Dissociative proton transfer reactions of H3+, N2H+, and H3O+ with acyclic, cyclic, and aromatic hydrocarbons and nitrogen compounds, and astrochemical implications

Article abstract of DOI:10.1021/jp014659i

A flowing afterglow-selected ion flow drift tube has been used to measure the rate coefficients and product ion distributions for reactions of H₃O⁺, N₂H⁺, and H₃⁺ with a series of 16 alkanes, alkenes, alkynes, and aromatic hydrocarbons as well as acrylonitrile, pyrrole, and pyridine. Exothermic proton transfer generally occurs close to the collision rate. The reactions of H₃O⁺ are mostly nondissociative and those of H₃⁺ are mostly dissociative, but many reactions, especially those of N₂H⁺, have both dissociative and nondissociative channels. The dissociative channels result mostly in H₂ and/or CH₄ loss in the small hydrocarbons and in toluene, loss of C₂H₂ from acrylonitrile, and loss of HCN from pyrrole. Only nondissociative proton transfer is observed with benzene, pyridine, and larger aromatics. Drift tube studies of N₂H⁺ reactions with propene and propyne showed that increased energy in the reactant ion enhances fragmentation. Some D₃⁺ reactions were also investigated and the results suggest that reactions of H₃⁺ with unsaturated hydrocarbons B proceed through proton transfer that forms excited (BH⁺)* intermediates. Pressure effects suggest that a fraction of the (BH⁺)* intermediates decomposes too rapidly to allow collisional stabilization in the flow tube (t < 3 × 10^-8 s). The other low-energy (BH⁺)* intermediates are formed by the removal of up to 40% of the reaction exothermicity as translational energy, and these intermediates result in stable BH⁺ products. The results suggest that, in hydrogen-dominated planetary and interstellar environments, the reactions of H₃⁺ can convert C₂-C₆ hydrocarbons to smaller and less saturated molecules, but polycyclic aromatics are stable against decomposition by this mechanism. The dissociative reactions of H₃⁺ can therefore favor the accumulation of small unsaturated hydrocarbons and aromatics in astrochemical environments.

Full text of DOI:10.1021/jp014659i

J. Phys. Chem. A 2002, 106, 9745-9755
9745
+
Dissociative Proton Transfer Reactions of H3 , N2H+, and H3O+ with Acyclic, Cyclic, and
Aromatic Hydrocarbons and Nitrogen Compounds, and Astrochemical Implications†
Daniel B. Milligan, Paul F. Wilson, Colin G. Freeman, Michael Meot-Ner (Mautner),* and
Murray J. McEwan*
Department of Chemistry, UniVersity of Canterbury, PriVate Bag 4800, Christchurch 8001, New Zealand
ReceiVed: December 31, 2001
A flowing afterglow-selected ion flow drift tube has been used to measure the rate coefficients and product
+
ion distributions for reactions of H3O+, N2H+, and H3 with a series of 16 alkanes, alkenes, alkynes, and
aromatic hydrocarbons as well as acrylonitrile, pyrrole, and pyridine. Exothermic proton transfer generally
+
occurs close to the collision rate. The reactions of H3O+ are mostly nondissociative and those of H3 are
mostly dissociative, but many reactions, especially those of N2H+, have both dissociative and nondissociative
channels. The dissociative channels result mostly in H2 and/or CH4 loss in the small hydrocarbons and in
toluene, loss of C2H2 from acrylonitrile, and loss of HCN from pyrrole. Only nondissociative proton transfer
is observed with benzene, pyridine, and larger aromatics. Drift tube studies of N2H+ reactions with propene
+
and propyne showed that increased energy in the reactant ion enhances fragmentation. Some D3 reactions
were also investigated and the results suggest that reactions of H3+ with unsaturated hydrocarbons B proceed
through proton transfer that forms excited (BH+)* intermediates. Pressure effects suggest that a fraction of
the (BH+)* intermediates decomposes too rapidly to allow collisional stabilization in the flow tube (t < 3 ×
10-8 s). The other low-energy (BH+)* intermediates are formed by the removal of up to 40% of the reaction
exothermicity as translational energy, and these intermediates result in stable BH+ products. The results suggest
that, in hydrogen-dominated planetary and interstellar environments, the reactions of H3+ can convert C2-C6
hydrocarbons to smaller and less saturated molecules, but polycyclic aromatics are stable against decomposition
by this mechanism. The dissociative reactions of H3+ can therefore favor the accumulation of small unsaturated
hydrocarbons and aromatics in astrochemical environments.
Introduction
exothermic and often dissociative process. In addition, such
exothermic proton transfer reactions are usually fast, proceeding
at collision efficiencies near unity.
The unique physical conditions in the interstellar medium
(ISM) of very low temperatures and number densities result in
ion-molecule reactions as a main route for the formation of
complex molecules.1 Modeling of the ISM has motivated many
studies of ion-molecule reactions, some by the selected ion
flow (SIFT) method.
Any build-up of large molecules in this medium must be
balanced against their dissociation to more simple species. This
is the reason that complex molecules in the ISM are observed
mainly in dark interstellar clouds, as the higher dust particle
concentrations in these clouds shield the molecules from
dissociation by photons and cosmic rays. These energy sources
can cause dissociation of the molecules directly, or the ions
can absorb some energy and will subsequently dissociate
complex molecules via highly exothermic reactions.
Some of the reactions studied in this work have been
+
investigated previously. Early studies of H3 reactivity were
done by Wexler,5 by Munson et al.,6 and by Aquilanti and
Volpi,7 where the last authors studied the reactions in a mass
spectrometer with a radioactive ion source. Their pioneering
work was done before the application of flow tubes to ion-
molecule studies, and many of the rate coefficients and product
assignments they recorded are questionable. The first flow tube
investigation of H3+ reactions was by Burt et al.8 in 1970, using
a flowing afterglow technique. They reported that H3+ reacted
with a number of simple molecules, mainly by proton transfer.
The reactions of H3+ with oxygen-containing organics were later
examined in some detail by Adams and co-workers.9 They
observed quite a number of product channels as a result of
dissociative proton-transfer reactions. Spanel et al.10 have
examined the proton-transfer reactions of several protonated
species with benzene, cyclopropane, and cyclohexane. Smith
et al.11 and Spanel et al.12 have also studied the proton-transfer
One of the prime candidates for this dissociative role is the
H3+ ion, which is formed readily when a hydrogen atmosphere
is ionized. The H3+ ion has been observed in interstellar space,
in both diffuse and dense molecular clouds,2 as well as in the
ionospheres of Jupiter, Saturn, and Uranus.3 Hydrogen is
+
reactions of H3 and H3O+ with a number of reagents.
+
abundant in these regions and as a result H3 will be an
Many dissociative charge and proton-transfer reactions pro-
duce both dissociation products and the undissociated molecular
ions.11-13 A basic question concerning these reactions is the
cause of this distribution; in particular, why undecomposed
products are formed when highly exothermic dissociation
channels are available. For a fuller understanding of these
dissociative reactions we need to consider the mechanism of
important protonating agent. Hydrogen has an extremely low
proton affinity (PA)4 and therefore proton transfer from proto-
nated hydrogen, i.e., H3+, to most other species is a highly
† Part of the special issue “Jack Beauchamp Festschrift”.
* Corresponding authors: m.mautner@chem.canterbury.ac.nz and m.
mcewan@chem.canterbury.ac.nz.
10.1021/jp014659i CCC: $22.00 © 2002 American Chemical Society
Published on Web 08/09/2002

9746 J. Phys. Chem. A, Vol. 106, No. 42, 2002
Milligan et al.
+
the reactions. Dissociative proton transfer (DPT) may proceed
first through sufficiently excited protonated intermediates
(BH+)*, which subsequently decompose to a fragment ion F+
and neutral or neutrals N. This reaction may occur in competi-
tion with the energy removal processes of collisional or radiative
stabilization:
afterglow (FA) source to generate H3 ions, which were then
injected into a helium carrier in the SIFT reaction tube. The
second method used helium with trace amounts of added krypton
mixed in with the helium in the FA tube. The helium/krypton
mixture was passed through the microwave discharge to generate
Kr+ ions, which were injected into a hydrogen carrier gas in
the SIFT tube. The Kr+ then reacts with the hydrogen carrier
+
gas to generate H3 ions.15 Hereafter the methods will be
denoted by the type of carrier gas in the SIFT reaction tube:
viz., as the helium or hydrogen method, respectively.
By the helium method, a helium/hydrogen plasma is first
generated in the FA source by a microwave discharge in the
helium gas. This plasma was created by passing helium through
a microwave discharge that generated a fraction of helium
metastable atoms and ions. These metastable atoms and ions
reacted subsequently with hydrogen added further downstream
AH+ + B fm (BH+)* + A
(1)
(2a)
(2b)
M, -hν
(BH+)*
m8 BH+
fm F+ + N
The dissociative reaction is therefore composed of a proton-
transfer reaction, for which the kinetics are well understood,
and unimolecular dissociation, where the theory is also well
established. However, the energy distribution in the population
of the (BH+)* intermediates is not as well characterized. In a
fraction of the reactions, some of the proton-transfer exother-
micity may be removed in inactive modes such as translation
so that the (BH+)* ions do not have enough energy to
decompose, and this fraction always yields the stable BH+
product. The other fraction of the (BH+)* ions will have enough
energy to decompose, but part of this population may be
stabilized collisionally, and the yield of BH+ vs F+ products
from this fraction should be pressure-dependent. The energy
effects can be investigated by varying the internal energy of
(BH+)* by proton-transfer reactions of varying exothermicity
+
in the ion source flow tube. Many of the H3 ions formed in
this plasma have vibrational excitation. Some of this excitation
will be quenched by collisions with the carrier gas before the
ions exit the FA source, but a fraction of the ions may retain
some vibrational excitation through the mass selection and
injection process. The H3+ ions exiting from the FA are mass-
selected by a quadrupole mass spectrometer and injected into
the helium carrier gas. The process of injection has a significant
effect on the energies of the H3+ ions. Some of these ions may
gain enough energy to overcome the endothermicity involved
in transferring a proton to the He carrier gas and thus generating
HeH+. With this in mind, the injection conditions were tuned
to the lowest practicable energy by minimizing the amount of
HeH+ formed in the injection process and also by titrating the
energy levels of the reactant ion with argon as the neutral
reactant. Proton transfer to both helium and argon is endothermic
with H3O+, N2H+, and H3 reagent ions or by exciting the
+
reagent ions by collisions in a drift tube. In the present work,
we shall use both methods to examine energy effects on the
product distributions.
from ground-state H3+, and only excited H3 will proton
+
Experimental Section
transfer:
The results were obtained by use of the University of
Canterbury flowing afterglow-selected ion flow-drift tube (FA-
SIFDT). The main components are a flowing afterglow tube,
where the reagent ions are generated by microwave discharge;
a quadrupole mass filter and venturi injector, where ions from
the FA are mass-selected and then injected into the reaction
zone in the flow tube; and the flow drift tube, where the ions
react with neutral molecules in a flowing He or H2 carrier gas.
The instrument has been described elsewhere.14 All reactions
are performed at room temperature (300 ( 5 K) and the ions
are thermalized at this temperature, except when electric fields
are applied in the drift tube. The pressure in the flow tube was
typically 0.480 Torr with a helium carrier gas (corresponding
flow rate ) 254 atm cm3 s-1 and flow velocity of 13 603 cm
s-1) and 0.340 Torr when hydrogen was used as the carrier gas
(corresponding flow rate ) 166 atm cm3 s-1 and flow velocity
of 12 359 cm s-1).
A major consideration in the present experiments is the
internal excitation of the reactant ions. The H3+ ions have large
vibrational quanta, and vibrational excitation can contribute
significant energy to the reactions. In addition, excited H3+ ions
can also react with the He carrier gas to generate HeH+ ions
that can contribute additional dissociative proton-transfer reac-
tions. The PA of helium is lower than that of hydrogen and
thus the reactions of HeH+ are expected to be more dissociative
than those of H3+. The presence of more energetic ion species
can significantly interfere with the determination of the products
from H3+ reactions. To check these effects, the H3+ ions in this
study were generated by two different methods that used either
helium or hydrogen carrier gas in the SIFT reaction tube. The
first method used a helium/hydrogen plasma in the flowing
H3+ (V * 0) + He fm HeH+ + H2
H3+ (V > 1) + Ar fm ArH+ + H2
(3)
(4)
Reaction 3 is endothermic by 245 kJ mol-1 and can occur
+
only when the H3 ions are very excited, possibly due to the
injection process itself. On the other hand, reaction 4 is
+
endothermic for ground-state H3 by only 53 kJ mol-1, which
can be overcome by excitation energy over any of the three
lowest vibrational levels of H3+. Accordingly proton transfer
+
to argon may occur unless the H3 is in the ground state or
either of the two V ) 1 vibrational modes (bending or breathing).
Consequently, when argon is added to the reaction flow tube,
the magnitude of the ArH+ signal gives a good indication of
+
the excitation levels of the H3 ions. This test was performed
several times and the fraction of excited H3+ ions was generally
less than 10% when the HeH+/H3 level was at a minimum
+
due to optimized injection conditions.
A similar method was used for generating the D3 reactant
+
ions used for reactions with some of the small hydrocarbons.
A moderately large flow of D2 was introduced into the helium
+
plasma at the FA nose cone and the D3 (generated via the
+
reaction of D2 with D2) was mass-selected and injected into
+
the SIFT reaction tube. One complication for a D3 reactant
ion is that D3+ and HeD+ ions (formed from energetic injection
of D3+) are isobaric. As the reactions of excited D3 and any
+
+
HeD+ with argon are exothermic, by comparing the D3 and
ArD+ ion counts at high argon flows, we found that the energetic
ions [HeD+ and D3+(V * 0)] comprised e3% of the 6 amu
signal.

Proton Transfer Reactions: Astrochemical Implications
J. Phys. Chem. A, Vol. 106, No. 42, 2002 9747
The second method for generating H3+ was to inject a signal
of Kr+ from the FA into a hydrogen carrier in the SIFT reaction
Reactions of H3O+ and N2H+ were included in this study for
the purpose of deconvolution and for investigating the effects
of reaction exothermicity.
+
tube, forming H3 through the reaction sequence 5 and 6:15
To obtain accurate product distributions, the product ion
counts generated by the ionic impurities were subtracted from
each product channel. In most cases this correction was simply
the branching ratio for the specific product channel multiplied
by the number of counts intially due to the H3O+ or N2H+ ion.
Where appropriate, a correction for mass discrimination was
also applied.
k ) 2.1 × 10-10 cm3
s
Kr+ + H2
-1m8 KrH+ + H + 27 kJ mol-1 (5)
k ) 3.8 × 10-11 cm3
s
KrH+ + H2
-1m8 H3+ + H2 - 2 kJ mol-1 (6)
The large number density of hydrogen molecules in the flow
tube ensures the complete conversion of KrH+ to H3+, despite
the low rate coefficient for this process. The Kr+ ions are
generated by direct microwave discharge of the helium carrier
gas in the FA to which a trace of Kr was added. This approach
has several advantages over the helium method. First, as the
H3+ is not injected into the SIFT tube, it cannot become excited
by the injection process. Second, hydrogen is expected to be
more effective at quenching any vibrationally excited H3+ ions
through symmetric proton transfer reactions. Finally, the forma-
tion reaction 6 is slightly endothermic and thus should produce
+
A special case occurs when C2H5 is a product ion in any
reaction. When this occurs, there is a mass overlap between
C2H5+ and the impurity ion N2H+ at m/z ) 29. This complicates
the estimation of the effect of contamination from N2H+
reactions. For these reactions, an estimate of the number of
counts of N2H+ to be subtracted was computed from the
measured flow of the neutral and the separately measured rate
coefficient for the reaction between that neutral and N2H+.
The measurement of the flow rates of some of the neutrals
added to the helium carrier gas in the SIFT tube was made
difficult by their low vapor pressures or by their adsorption in
the flow lines. In the case of naphthalene and methylnaphtha-
lenes, the rate coefficients could not be determined due to these
complications and only the products are reported to an accuracy
of about 10%. With acrylonitrile, a mixture of acrylonitrile vapor
in helium was required to prevent problems with adsorption of
acrylonitrile to the system walls. The composition of this mixture
of ∼10% acrylonitrile was calibrated by assuming that the
reaction between H3O+ and acrylonitrile proceeded at the
collision rate (5.1 × 10-9 cm3 s-1).17 Since some of the reactions
only ground-state H3+. This second method of generating H3
should therefore effectively eliminate any vibrationally excited
+
+
H3 ions. However, this cannot be verified experimentally by
an argon titration of excited H3 ions, as the resulting ArH+
+
+
ions will react with the hydrogen carrier to regenerate H3
.
The large number density of hydrogen molecules in the SIFT
reaction tube can, however, create complications in the deter-
mination of the ionic products. Even slow association reactions
with hydrogen (rate coefficients larger than 5 × 10-31 cm6 s-1
)
can have significant effects on the observed product distribution.
Association reactions with hydrogen are especially problematical
as they essentially reverse the dissociation reactions that involve
H2 loss. In these cases the product distributions in the hydrogen
carrier gas will underestimate the amount of dissociation. For
+
of H3 reported in this study proceed more slowly than the
collision rate, this assumption may slightly overestimate the
reported rate coefficients for acrylonitrile reactions.
+
all these reasons, a combination of the two H3 generation
Results
methods was required to determine accurately the product ratio
produced by ground-state H3+ ions. Some small differences in
product distribution are to be expected to arise from the different
diffusion rates of the product ions in helium as compared with
hydrogen. However, the substantial differences reported here
in a few reactions are not due to differential diffusion.
There is one additional complication arising from trace neutral
contaminants (particularly N2 and H2O) present in the carrier
gas, as these two contaminants could not be completely
The reactions of H3+, N2H+, and H3O+ have been investigated
with a wide range of hydrocarbons as well as pyridine, pyrrole,
+
and acrylonitrile. Most of the reactions of H3 have been
investigated with both helium and hydrogen carrier gases.
The rate coefficients for the reactions of H3+, N2H+, and
H3O+ with the range of neutrals investigated are shown in Table
1, along with the theoretical collision rate coefficients.17 The
+
rate coefficients for the reactions of H3 are considered to be
+
eliminated. Because of the rapid reactions between H3 and
nitrogen (reaction 7) and between H3 and water (reaction 8)
slightly less accurate than most measured on the SIFT due to
the increased uncertainties in measuring the small flows required
for accurate rate coefficient determination in these fast reactions.
+
and the similarly fast reaction between N2H+ and water (reaction
9),16 even trace levels of these neutral contaminants will cause
significant impurity ion signals at m/z ) 29 and m/z ) 19 to be
observed. Typical water levels in helium in the flow tube after
passage of the helium through a liquid nitrogen-cooled zeolite
trap are about 150 parts per billion.
+
The uncertainty of H3 rate coefficients is assigned as (20%
rather than the usual uncertainties of (15% in SIFT measure-
ments that apply here to the H3O+ and N2H+ rate coefficients.
The products observed for the three different ionic reactants,
including the two different methods of H3+ formation, are shown
in Table 2.
The ions thus formed in reactions 7 and 9 will react with the
added neutral reactants via proton transfer and yield product
distributions different than the reactions of H3 alone. For a
particular neutral reactant, these products are included in the
detected product distribution and the products need to be
deconvoluted to find the yields of the H3+ reactions themselves.
+
Discussion
(1) Alkanes. Five alkanes have been studied: two linear
(propane and n-butane), one branched (iso-butane) and two
cyclic (cyclopropane and methylcyclohexane).
(a) H3O+ Reactions. All the reactions of H3O+ with the
acyclic alkanes have low rate coefficients, whereas the cycloal-
kane reactions are fast. The small rate coefficients for acyclic
alkanes are due to the fact that their low proton affinities (PAs)
make proton transfer slightly endothermic. The rate coefficient
for propane is below the lowest limit that can be measured by
k ) 1.9 × 10-9 cm3
s
H3+ + N2
H3+ + H2O
N2H+ + H2O
-1m8 N2H+ + H2
(7)
(8)
(9)
k ) 5.3 × 10-9 cm3
s
-1m8 H3O+ + H2
k ) 2.6 × 10-9 cm3
s
-1m8 H3O+ + N2

9748 J. Phys. Chem. A, Vol. 106, No. 42, 2002
Milligan et al.
+
TABLE 1: Rate Coefficients Measured in This Work
Compared with Calculated Collision Rate Coefficients from
Ref 17 and Previous Measurements
produce C2H5 + CH4, where the product ion is isobaric with
the N2H+ reactant ion. For this reason, we believe the true rate
coefficient for the reaction with propane is close to the collision
ionic reactant
rate coefficient of 1.4 × 10-9 cm3 s-1
.
+
H3O+
N2H+
H3
+
(c) H3 Reactions. The rate coefficients for the reactions of
neutral
this work ref 17 this work ref 17 this work ref 17
H3 with the alkanes are all approximately 3 × 10-9 cm3 s-1
,
+
acetylene
0.01
1.3
1.4 (1.4a) 1.2 2.6 (2.9,a
1.9,b 3.5c)
1.3 2.7 (2.0,b
3.6,c 2.8f)
2.8
2.9
being marginally lower than the respective collision rates. The
reaction with propane is relatively simple, and only the two
dissociative channels with exothermicities greater than ∼50 kJ
ethylene
0.06 (0.06d) 1.4
e
allene
1.4
1.8
1.4
2.0
1.5
1.7
1.4
1.5
1.7 2.9
1.7 3.0
3.4
4.2
3.3
3.7
3.5
11.3
4.2
4.0
4.0
6.4
7.9
4.4
5.8
5.2
mol-1 (C3H7+ and C2H5+) are observed. The formation of C3H5
+
propyne
cyclopropane 1.6 (1.5g)
1.4 (1.3h) 1.3 2.6 (3.0h)
is exothermic by 57 kJ mol-1, but this product is not observed.
The formation of this product would require the sequential loss
of two H2 molecules, and the first dissociation step may remove
enough energy, preventing further dissociation of the C3H7+ ion.
The reactions of H3+ with the two butanes gave predominantly
the ions C4H9+, C3H7+, and C2H5+. For n-butane, the exothermic
channels are shown in reaction 10. In calculating the thermo-
chemistry, we assume that the unbranched structure has been
retained.
propene
propane
1.7 (1.5i)
NRj
1.4
1.5 3.1
1.4 3.0
4.4 9.1
1.6 2.8
1.5 3.0
1.5 3.0
2.5 4.9
2.9 3.5
1.6 >0.63k
acrylonitrilel 5.1
5.1
1.9
1.8
1.8
2.9
3.4
1.9
2.2
2.2
4.4
1.3
1.3
0.98
2.2
2.3
2-butene
isobutane
n-butane
pyrrole
pyridine
benzene
toluene
1.6
<0.004
<0.003
2.4
2.5
1.3 (1.8g)
1.5 (1.6h) 1.6 3.3 (3.9h)
1.3
1.7
1.3
1.8 3.9
1.9 4.2
methylcyclo- 0.71
hexane
a Reference 18. b Reference 8. c Reference 19. d Reference 20. e The
rate coefficient for this reaction could not be measured because of mass
coincidence. On the basis of the difference in PAs, it should occur at
close to the collision rate. f Reference 21. g Reference 12. h Reference
H3+ + n-C4H10
fm CH3CHCH2CH3+ + 2H2 + 271 kJ mol-1
(10a)
10. i Reference 22. j NR ) no reaction; k < 5 × 10-13 cm3 s-1
.
k The
fm CH3CHCH3+ + CH4 + H2 + 257 kJ mol-1 (10b)
rate coefficient for this reaction is believed to be close to the collision
rate of 1.4 × 10-9 cm3 s-1; see text. l The rate coefficients for
acrylonitrile are relative rate coefficients based on the collision rate
being assumed for H3O+/acrylonitrile.
fm C3H5+ + CH4 + 2H2 + 110 kJ mol-1
fm C2H5+ + C2H6 + H2 + 163 kJ mol-1
(10c)
(10d)
the SIFT technique, meaning that propane is essentially unre-
active. An ionic product was observed at mass 43 (C3H7+) but
this is almost certainly due to a trace impurity, probably propene,
The yields of the product ions generally follow the exother-
micities, with C3H7+, C4H9+, and C2H5+ being the main product
ions observed. The other possible product ion, C3H5+, is
observed only as a minor product in the helium carrier gas,
+
in the reagent gas. The lowest energy C3H7+ isomer (c-C3H7
)
possible in the reaction of H3O+ with C3H8 is 65 kJ mol-1
endoergic. The rate coefficients for n-butane and iso-butane were
much higher than for propane and fall within the measurable
range for the SIFT, although Spanel and Smith24 did not report
a reaction with n-butane. In both reactions, proton transfer is
dissociative and produces C4H9+, as the C4H11+ ion that would
correspond to nondissociative proton transfer is unstable. In
addition, both reactions also yield the association product
(H3O+‚C4H10). Spanel and Smith12b noted this association of
H3O+ with n-alkanes for n-pentane and larger homologues. The
ion formed by the protonation of methylcyclohexane (C7H14),
i.e., C7H15+, must have a small barrier to further dissociation,
as it is formed only when the difference in PAs between the
proton acceptor and proton donor is small. The loss of one
hydrogen molecule, C7H13+, is more prevalent as an outcome
from proton transfer. The rate coefficient for this process is
higher than for either butane isomer and the product distribution
we observe is in good agreement with the earlier measurement.12
(b) N2H+ Reactions. The reactions of N2H+ with the alkanes
are more rapid than those of H3O+, and they become increas-
ingly dissociative as the complexity of the neutral species
increases. Propane appears to react at just under half the collision
rate to give solely a C3H7+ product. This process is exothermic
+
where it may be due to excited H3 ions. With i-C4H10 the
relative energetics and product distributions are similar, and the
+
yield of C2H5 formed is smaller than with n-C4H10, possibly
because isomerization of i-C4H10 is required to form this product.
(2) Alkenes and Cycloalkanes.
(a) H3O+ Reactions. In the reactions of H3O+ with the
alkenes, the smallest alkene, ethylene, reacts at about 4% of
the collision rate, since this proton transfer is endothermic by
11 kJ mol-1, which also accounts for the competitiveness of
the association channel. Proton transfer to the other alkenes is
exothermic and proceeds near the collision rate, as was also
noted by Spanel et al.12 A fast reaction is also observed with
cyclopropane, where C3H7+ is the only product. This behavior
is similar to the reactions of H3O+ with alkenes and may involve
the ring opening of the neutral reactant in the reaction complex
so that the reaction effectively involves the interaction of H3O+
with propene.
The only dissociative reaction of H3O+ observed in this work
is in the reaction with methylcyclohexane. This is also the least
exothermic dissociative reaction observed in this study. Spanel
et al.12 show a similar product distribution for this reaction.
by 51 kJ mol-1 to form n-C3H7 or 133 kJ mol-1 to form iso-
+
C3H7+, and the latter is the likely product. These energies and
the fact that the proton affinity of propane is higher by 133 kJ
mol-1 than that of nitrogen make the somewhat low reaction
efficiency unexpected, as exothermic proton transfer usually
proceeds at the collision rate.25 It is almost certain that a
significant fraction of reactions (>0.3) are dissociative and
(b) N2H+ Reactions. Dissociative channels are generally
observed when N2H+ is the reactant ion, although the reaction

Proton Transfer Reactions: Astrochemical Implications
TABLE 2: Product Distributions in the Reactions of H3O+, N2H+, and H3 with Neutral Moleculesa
H3O+ N2H+ H3+ (H2 carrier)
J. Phys. Chem. A, Vol. 106, No. 42, 2002 9749
+
H3+ (He carrier)
product -∆H° b product -∆H° b product -∆H° b product -∆H° b
neutral
productsa
ratio
(kJ/mol)
ratio
(kJ/mol)
ratio
(kJ/mol)
ratio
(kJ/mol)
Alkanes
-65
-93
-5
C3H8
C3H7+ + H2
NR
1.00
133
-104
192
178
84
199
185
0.40
0.60
0.35
0.60
>0.05
0.35
0.45
0.55
0.30
0.50
0.20
0.35
0.45
0.05
0.15
204
176
264
250
155
271
257
110
163
C2H5+ + CH4
C4H9+ + H2
iso-C4H10
0.60c
0.20
0.80
C3H7+ + CH4
C2H5+ + C2H6
C4H9+ + H2
-19
-113
2
n-C4H10
0.70c
0.25
0.75
C3H7+ + CH4
C3H5+ + CH4 + H2
C2H5+ + C2H6
-12
0.45
-106
91
0.20
Alkenes and Cycloalkanes
+
C2H4
C2H5
0.75d
-11
-224
84
1.00
+187
-27
281
152
48
258
110
89
-12
-18
-122
16
291
257
144
122
21
1.00e
0.40
0.60
259
45
C2H3+ + H2
CH2dCdCH2
CH2CHCH2+ (allyl)
C3H3+ (cyclic) + H2
C3H3+ (propargyl) + H2
C3H7+ (isopropyl)
C3H5+ (allyl) + H2
C3H5+ (2-propenyl) + H2
C3H5+ (cyclopropyl) + H2
C3H3+ (cyclic) + 2H2
C3H3+ (propargyl) + 2H2
C2H3+ + CH4
1.0
1.0
353
224
120
329
183
161
59
53
-51
88
362
328
215
194
92
-45
-150
61
-86
-108
-209
-216
-320
-181
94
1.00f
0.70
0.3
0.7
CH3CHdCH2
1.0
1.0
0.55
0.45
0.65
0.10
0.25
0.30g
0.25
c-C3H6
C3H7+ (isopropyl)
(protonated cyclopropane)
C3H5+ (allyl) + H2
C3H5+ (2-propenyl) + H2
C3H5+ (cyclopropyl) + H2
C3H3+ (cyclic) + 2H2
C3H3+ (propargyl) + 2H2
C2H5+ + CH2
0.10
0.90
59
-53
-75
-177
-183
-287
-396
-148
118
-17
-43
-126
59
0.60
15
86
-90
-199
49
315
180
155
71
257
290
237
33
-18
-127
121
387
252
226
142
328
362
309
104
C2H3++ + CH4
0.75
0
0.4
0.35
0.25
1.00
0.75
0.20
0.05
0.40
0
0.4
0.35
0.25
1.00
0.40
0.50
0.10
2-butene (C4H8)
C4H9
1.00
0.1
0.4
0.5
C4H7+ + H2
C3H5+ + CH4
C2H5++ + C2H4
benzene
toluene (C7H8)
C6H7+
1.00
1.00
1.00
1.00
C7H9
93
C7H7+ + H2
40
C6H5+ + CH4
-164
+
methylcyclohexane (C7H14) C7H15
0.10
0.90
C7H13+ + H2
29
50
0.40
0.30
0.30
226
247
151
141
101
70
309
341
340
0.60
0.25
0.15
0.35
0.20
0.25
0.10
0.10
298
319
222
213
172
142
381
413
412
C6H11+ + CH4
C4H9+ + C3H6
C4H7+ + C3H8
C3H7+ + C4H8
C3H5+ + C4H10
-46
-56
-97
-127
112
144
143
+
naphthalene
1-methylnaphthalene
2-methylnaphthalene
C10H9
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
+
C11H11
C11H11
+
Alkynes
-50
80
1.00e
1.0g
1.00
219
349
327
219
115
+
C2H2
CH3CtCH
C2H3
NR
1.0
1.0
1.0
148
277
255
148
44
C3H5+ (allyl)
C3H5+ (propenyl)
C3H3+ (cyclic) + H2
C3H3+ (propargyl) + H2
58
-50
-153
0.3
0.7
Nonhydrocarbons
CH2CHCN
pyrrole
CH2CHCNH+
HCNH+ + C2H2
C4H5NH+
1.00
94
-160
184
1.00
1.00
1.00
291
37
382
64
0.90
0.10
0.75
0.25
1.00
362
109
453
135
508
121
1.00
1.00
C3H5+ + HCN
C5H5NH+
-133
pyridine
239
436
C4H5+ + HCN
a Only the products formed from the reactant neutrals are shown; the deprotonated neutrals from the reagent ions, H2O, N2 and H2, are omitted
for clarity. b Exothermicity of the reaction channel, corresponding to the most stable ion and neutral isomers as products, unless otherwise indicated.
Thermochemical data are from ref 23. c The remainder of the observed product was C4H10‚H3O+, which was formed via an association-type reaction.
d The remainder of the observed product was C2H4‚H3O+, which was formed via an association-type reaction. e In H2 carrier gas C2H3+ was observed
to associate with H2, yielding C2H5+. The product of the H3+ + C2H2 reaction in H2 carrier was assigned after accounting for this reaction (see text).
+
+
+
The product observed in H3 + C2H4 in H2 carrier was C2H5+, which may originate in part from a C2H3 product associating with H2. f A C3H5
+
product was also observed but was attributed to the association of C3H3 with the H2 carrier gas, on the basis of observations in He carrier. g A
C2H5+ product was also observed but was attributed to the association of C2H3+ with the H2 carrier gas, on the basis of observations in He carrier.

9750 J. Phys. Chem. A, Vol. 106, No. 42, 2002
Milligan et al.
with ethylene is nondissociative, as hydrogen loss is endothermic
by 27 kJ mol-1. Dissociative proton transfer to allene to form
+
C3H3+ + H2 is exothermic by 48 kJ mol-1 to form linear C3H3
,
or by 152 kJ mol-1 for c-C3H3+, but is not observed. The fact
that the latter is not observed suggests that cyclization to form
the c-C3H3+ ion involves a significant barrier. The relationship
between energy deposition and product distributions will be
discussed later.
In 2-butene, H2 loss and CH4 loss have comparable energies
and similar yields. Although both are significantly exothermic,
the nondissociative channel is still observed, indicating that in
10% of the collisions, over 180 kJ mol-1 of the exothermicity
is removed in nonreactive modes (see later discussion).
The reactions of N2H+ with cyclopropane and methylcyclo-
hexane proceed near the collision rate and in both cases the
dissociative channels dominate. We note that the dissociative
channel is significantly more dominant in the reaction of N2H+
with cyclopropane than with its linear isomer, propene. This
may be due to a lower barrier to H2 loss from the cyclic isomer.
Alternatively, protonation may result in ring opening, leading
to the same lowest energy [(CH3)2CH+]* intermediate but with
more internal energy in the more exothermic c-C3H6 reaction,
which leads to increased dissociation. Note that H2 loss from
the (CH3)2CH+ ion may require a constrained four-center
intermediate and possibly a significant barrier. The results for
cyclopropane are in good agreement with those of Spanel et
al.10 Proton transfer to methylcyclohexane is also fast, possibly
assisted by the availability of four dissociative channels and
their corresponding transition states.
Figure 1. Observed product distribution for the reaction of 2-butene
with H3O+, N2H+, and H3+. The proton affinity difference PA(C4H8 -
X), where X ) H2O, N2, and H2, is used as an indication of the energy
contained in each ion. The curves show an interpolation between the
available data points.
Similar arguments to those made in the reaction of N2H+ with
allene discussed above, where the exothermic reaction to
produce c-C3H3+ is not observed, suggest that there is an energy
barrier for rearrangement to the cyclic ion.
The reactions of both propene and 2-butene are relatively
simple, giving the expected exothermic dissociative products.
+
Propene reacts to give mostly C3H5+ and C2H3+. Some C3H3
was seen in the helium carrier but not in the hydrogen carrier,
suggesting that it is produced by excited H3+ ions. Some of the
C3H5+ ions produced by the energetic reagent ions seem to retain
enough energy to also lose H2, although the overall process to
form C3H3+ by the reaction of ground-state H3+ is endothermic.
+
+
(c) H3 Reactions. The reaction of H3 with ethylene is
interesting in that the nondissociative channel is exothermic by
259 kJ mol-1 while the dissociative channel to yield C2H3 is
+
The final alkene studied was 2-butene, which gave only
+
dissociative ion products in reaction with H3+, viz., C4H7
exothermic only by 45 kJ mol-1, yet it is the major channel. In
other words, the energetics means that H2 loss from a ground-
state C2H5+ ion is endothermic by 214 kJ mol-1, and this much
energy must remain deposited in the internal modes of the
(C2H5+)* ion generated by proton transfer, to undergo the
subsequent dissociation (see later discussion). Fiaux et al.21
investigated this reaction with respect to the internal energy of
+
+
(0.40), C3H5 (0.35), and C2H5 (0.25). The only other
+
dissociations possible energetically are those to give c-C3H3
+ CH4 + H2, exothermic by 97 kJ mol-1, and C2H3+ + C2H6,
exothermic by 65 kJ mol-1. Both of these channels are only
moderately exothermic and as such do not compete well with
the observed channels that are exothermic by more than 200 kJ
mol-1. Formation of the cyclopropenyl ion is also apparently
unfavorable due to the barrier discussed above. The change in
product distributions from H3O+, N2H+, and H3+ with 2-butene
illustrates the shift toward more dissociative processes as the
+
+
the H3 ions and showed that the amount of C2H3 observed
increases with the internal energy of the H3+ ion. Interestingly,
this is accompanied by a much more rapid decrease in the C2H5
+
ion product as direct charge transfer (CT) to form C2H4+ begins
to occur. Charge transfer is 132 kJ mol-1 endothermic for
ground-state H3+. This observation indicates the amount of
energy that it is possible to hold in vibrational excitation of
H3+ and also shows that the ions used in this study are relatively
energy of the reactant ion increases as shown in Figure 1. Only
+
C4H9 is observed from H3O+ [∆PA(C4H8 H2O) ) 56 kJ
-
mol-1], whereas three dissociative processes are found from
H3 [∆PA(C4H8 - H2) ) 325 kJ mol-1].
+
thermalized as no CT products were seen. The ground-state
With cyclopropane, as with propene, only dissociative
products are observed. With both cyclopropane and propene in
the helium carrier gas, the product ions C3H3+, C3H5+, and
C2H3 are found, while in hydrogen only C3H5 and C2H3
are produced; the latter is observed as C2H5+ due to association
+ 21
product distributions of Fiaux et al. (70% C2H3+, 30% C2H5
)
also agree well with those obtained in the current study (60%
C2H3+, 40% C2H5+). Note that this branching ratio could be
obtained only from the helium carrier gas; C2H5+ was the only
product observed in a hydrogen carrier due to the association
+
+
+
+
+
of C2H3 with the H2 carrier gas. The C3H3 ion is a minor
product and is likely to be the result of small amounts of more
energetic ions in the helium carrier gas. Spanel et al.10 reported
+
reaction of C2H3 with hydrogen.
The reaction with allene is relatively simple and forms only
+
+
+
60% C3H5 and 40% C2H3 as the products of reaction with
H3 with cyclopropane, which is in good agreement with our
the dissociative product. The structure of the C3H3 product
+
ion was investigated through its reactions with methanol. The
reaction yielded a curved semilogarithmic decay of C3H3+ ion
signal with increasing CH3OH flow, indicating that two different
C3H3+ species with different reactivities were present. A model
using the rate coefficients of McEwan et al.26 found that
approximately 70% of these ions were the linear isomer and
only 30% were the more energetically favorable cyclic one.
findings in a helium carrier gas. By examining the dissociation
products of eight proton donors, HX, reacting with cyclopropane,
they also demonstrated that the dissociative product, C2H3+, was
produced only when the difference in PAs, ∆PA ) PA(c-C3H6)
- PA(HX), was greater than 292 kJ mol-1. Their finding is
consistent with our observations for 2-butene. As with N2H+,

Proton Transfer Reactions: Astrochemical Implications
J. Phys. Chem. A, Vol. 106, No. 42, 2002 9751
the reaction of H3+ with methylcyclohexane leads to extensive
fragmentation. It is also interesting to note that the distribution
(4) Aromatic Hydrocarbons. The aromatic hydrocarbons
studied were benzene, toluene, naphthalene, and 1- and 2-
methylnaphthalene. Because of their low vapor pressures, the
latter three were only studied in a dilute mixture in a helium
carrier gas and rate coefficients were not obtained, although it
is anticipated that proton transfer or dissociative proton transfer
will occur at close to the collision rate.
+
of product ions is similar with N2H+ and H3 despite the
difference between the reaction exothermicities. We are unable
to account for differences in the product ion distributions in
+
the H3 reaction with methylcyclohexane between the helium
and hydrogen carriers.
The reactions of N2H+ and H3 with cyclohexane were
+
The reactions of H3O+ and N2H+ with all the aromatics gave
only nondissociative proton transfer and had rate coefficients
that were close to the collision rate, in good agreement with
investigated by Spanel et al.,10 who found the major channels
to be C6H11+ (H2 loss) and C3H7+ (C3H6 loss). There was also
+
a shift to increased dissociative products for H3 with respect
+
the findings of Spanel et al.10,12 The reactions of H3 with all
to N2H+ that is similar to the present observations with
methylcyclohexane.
the aromatics, except toluene, yielded only the protonated parent
ions. Since these molecules have large polarizabilities, the
calculated rate coefficients are high but the experimental rate
coefficients are somewhat smaller.
(3) Alkynes.
(a) H3O+ Reactions. The reactions of H3O+ with the alkynes
follow a similar trend to those of the alkenes. Proton transfer
from H3O+ to acetylene is endothermic and only a slow
association reaction is observed, while exothermic proton
transfer to propyne undergoes proton transfer near the collision
rate.
The reactions of toluene are interesting. The reaction is
nondissociative with N2H+, although H2 loss would be exo-
thermic by 237 kJ mol-1 to form the tropylium ion or by 187
kJ mol-1 to form the methylenebenzyl cation. Increasing the
+
exothermicity by 72 kJ mol-1 in going to H3 leads to a
(b) N2H+ Reactions. Proton transfer from N2H+ to the alkynes
is exothermic, and nondissociative proton transfer proceeds near
the collision rate. In the case of acetylene this is the only possible
exothermic reaction, but with propyne an additional exothermic
dissociative channel is also possible.
significant H2 loss channel and a smaller CH4 loss channel. The
+
exact extent of this dissociation is difficult to gauge in the H3
/
toluene reaction, however, as the product ratios in the helium
and hydrogen carrier are quite different. In the helium carrier
the observed products were C7H9+ (0.40), C7H7+ + H2 (0.50),
+
and C6H5 + CH4 (0.10), whereas in the hydrogen carrier the
N2H+ + CH3CtCH
ratios were 0.75, 0.20, and 0.05. There are two possible reasons
+
fm CH2CHCH2+ + N2 + 277 kJ mol-1
(11a)
(11b)
for this observed difference in branching ratio: either the H3
in the helium carrier was vibrationally excited and amplified
fm l-C3H3+ + N2 + H2 + 44 kJ mol-1
+
the hydrogen loss channel, or C7H7 reacts with hydrogen
(possibly very slowly) to generate C7H9+. The product ion due
to methane loss observed in the hydrogen bath gas was actually
fm c-C3H3+ + N2 + H2 + 148 kJ mol-1 (11c)
+
C6H7+. However, the C6H5 ion cannot be observed in the
The ion in reaction 11a has been assigned the allyl structure,
as Fairley et al.27 have shown that when enough energy is
present to overcome the ∼110 kJ mol-1 barrier to interconver-
sion between the 2-propenyl and allyl structures, the allyl
presence of a hydrogen carrier, as Ausloos et al.,28 Giles et al.,29
and Scott et al.30 have shown that C6H5+ associates rapidly with
+
hydrogen to give C6H7+. Direct formation of C6H7 requires
either the energetically unfavorable elimination of CH2 (162
kJ mol-1 endothermic) or the unlikely insertion of the hydrogen
molecule into the ionic reactant followed by loss of CH4. We
+
structure is the only product. Even though formation of c-C3H3
in reaction 11c would be exothermic by 148 kJ mol-1, the
reaction is not observed. This observation is consistent with the
+
have therefore assigned the products of the reaction with H3
+
as C6H5+, which then rapidly associates with hydrogen to give
the observed C6H7+. Several groups have demonstrated that the
open or cyclic isomers of C6H5 react quite differently with
hydrogen, with the reactions of the cyclic ion being faster.29,30
The C6H5 ions observed in this study were completely
converted to C6H7 by the end of the flow tube, and therefore
we tentatively assign the cyclic structure to the C6H5 ions
evidence noted earlier for a barrier to the formation of c-C3H3
ions.
+
+
+
(c) H3 Reactions. The reactions of H3 with alkynes are
relatively simple. Proton transfer is exothermic in both cases.
Acetylene gives only the exothermic product, C2H3+; and
propyne, only the dissociative product. A small amount of
energetically impossible C3H+ was observed also in helium
carrier gas, apparently from a small HeH+ impurity ion and/or
+
+
+
formed in this reaction.
excited H3+. The C3H5 cation seen in the H2 carrier gas may
be formed by the association of C3H3 with H2. For propyne
the C3H3 ion was, as in the allene case above, formed in two
+
(5) Nitrogen Compounds. Three neutrals containing nitrogen
were also investigated: pyridine, pyrrole, and acrylonitrile. Of
these molecules, only acrylonitrile, CH2CHCN, has been
observed in the interstellar medium at this time.31 The H3O+
and N2H+ ions reacted with all three species by simple proton
transfer. The rate coefficients obtained for acrylonitrile were
measured by use of a mixture of the vapor in helium, where
the concentration of acrylonitrile was calibrated relative to the
rate coefficient observed from H3O+. The reactions with N2H+
also all proceeded via nondissociative proton transfer with rate
coefficients near the collision rate.
+
+
isomeric structures. Approximately 70% of the products were
+
l-C3H3 and 30% were c-C3H3+, and again the formation of
the energetically more favorable cyclic ion as the smaller product
suggests a barrier.
An early investigation of this reaction by Aquilanti and Volpi7
+
+
+
saw a mixture of C3H3 and C3H5 ions, with more C3H5
forming as the pressure was increased. Their observation is
+
consistent with association of the C3H3 product ion with
+
hydrogen. The fact that no C3H5 is observed in the helium
The reactions with H3+ were more complex. Nondissociative
proton transfer was still the only product observed for pyridine,
but dissociative channels were observed with pyrrole and
carrier gas and that it apparently represents only ∼30% of the
+
products in the hydrogen carrier strongly suggests that C3H5
is not a primary product in this reaction.

9752 J. Phys. Chem. A, Vol. 106, No. 42, 2002
Milligan et al.
TABLE 3: Ratios of Observed vs Statistical HD/H2 Loss or
CH3D/CH4 Lossa from (CXHYD+)* Formed from the
acrylonitrile.
+
Reaction of D3 with a CXHY Hydrocarbon
H3+ + C4H5N
hydrogen/deuterium ratio
neutral
product
ion
fm C4H5NH+ + H2 + 453 kJ mol-1 (0.75)
(12a)
hydrocarbon
experimental
statistical
+
fm C3H5+ + HCN + H2 + 135 kJ mol-1 (0.25) (12b)
ethylene
propene
C2H3
C3H5
C2H3
C3H3
C2H5
C3H7
0.50
0.44
1.30
0.57
3.0
0.67
0.40
1.32
0.67
0.80
0.28
+
+
+
H3+ + CH2CHCN
fm CH2CHCNH+ + H2 + 362 kJ mol-1 (0.90)
propyne
propane
+
+
<4.0
(13a)
a Equivalently, D loss vs D retention.
fm HCNH+ + C2H2 + H2 + 109 kJ mol-1 (0.10) (13b)
A special case in this work is that of propane, where the ion
formed upon initial protonation, C3H9+, is not a stable species.
In this reaction there is essentially no isotope scrambling and
almost no deuterium is retained in the products. The results
suggest a direct mechanism. Note that hydride abstraction
constitutes attack of the incoming proton on a hydrocarbon
hydrogen, while the fast dissociation mechanism involves
protonation of a carbon atom or of a C-H bond followed by
rapid H2 loss. Since the C-H bonds in propane are similar to
those in the alkyl groups of the other hydrocarbons and the
hydrogen atoms carry similar partial charges, it is unlikely that
hydride abstraction is easier here than in the other hydrocarbons
and the rapid dissociation mechanism is more likely. This is
also consistent with the observations of Futrell and co-
workers21,32-35 of increased direct reaction products with
Reaction 12b requires considerable rearrangement and loss
of the aromaticity of pyrrole but is still significantly exothermic.
A similar channel for pyridine leading to C4H5+ + HCN + H2
is 121 kJ mol-1 exothermic but it is not observed, possibly due
to a barrier resulting from the significant aromaticity of the
molecule.
(6) Reaction Mechanisms. The mechanisms of dissociative
proton-transfer reactions, including those of H3+, were inves-
tigated extensively by Futrell and co-workers21,32-35 and more
recently by Spanel et al.10 The discussion below follows similar
arguments.
The loss of H2 can occur by two basically different mecha-
nisms. One mechanism involves proton transfer to form an
excited protonated intermediate, followed by the dissociation
of this intermediate. If the intermediate is sufficiently long-lived,
scrambling can occur among the hydrogen atoms in the
intermediate, which can be tested by deuterium labeling. The
second mechanism involves direct reaction by hydride abstrac-
tion or protonation of a carbon atom followed by rapid loss of
the incoming proton without scrambling. The hydride abstraction
or rapid loss mechanisms are indistinguishable by deuterium
labeling. These reactions are shown below.
increasing ion energy, which also facilitates rapid dissociation.
Spanel et al.10 also found a nonstatistical distribution of C3H5
/
+
C3H4D+ in the reaction of ArD+ with cyclopropane and
concluded that the more direct process of dissociative hydride
ion transfer, rather than proton transfer, was responsible.
Another question about the mechanism is whether collisional
stabilization, reaction 2a above, is significant. Information on
this point can be obtained by pressure effects, especially from
a comparison with the results of Smith and Futrell.21,32-35 The
product ratios in their experiments, obtained at low pressures
(<10-5 Torr), are very similar to the results obtained in the
SIFT. Thus one can conclude that the lifetime of the excited
ions, i.e., the time required for unimolecular dissociation, is less
than the time between collisions in the SIFT at 0.5 Torr (∼3 ×
10-8 s). The presence of the carrier gas does not appear to
significantly alter the product distribution. This suggests that
reaction 2a above is not significant for the reactions observed
both here and in the ICR experiments.
Proton transfer:
XH+ + CxHy fm (CxHy+1+)* + X
(CxHy+1+)* fm CxHy-1+ + H2
(14)
(15)
Hydride abstraction or rapid loss:
XH+ + CxHy fm CxHy-1+ + H2 + X
(16)
+
In the present work we examined the reactions of D3 with
several of the small hydrocarbons. First, the results showed an
almost complete lack of any product ions containing two
deuterium atoms. This finding is consistent with practically all
other investigations that have attempted isotopic labeling and
is independent of whether the labeling was present in the reactant
(7) Energy Effects. The main energy effect observed in the
present reactions is the increasing dissociation with exother-
micity in the reactions of H3O+, N2H+, and H3 with a given
+
neutral. To investigate further the role that energy plays in these
reaction systems, we examined the reactions of N2H+ with
propene and propyne in the drift tube. The product ratios at
different collision energies were obtained in a 0.34 Torr helium
carrier gas with drift tube voltages ranging from 0 to -275 V.
The voltage of -275 V is the maximum voltage applied over
50 rings uniformly spaced over 50 cm with a constant potential
gradient between each ring. The graph obtained for propene is
shown in Figure 2.
+
ion or the neutral.10,21,32-35 In other words, if H3+/D3 was
reacted with CxDy/CxHy, at most one of the hydrogen isotopes
from the reactant ion was seen in the ionic products. This
observation suggests that proton transfer proceeds though long-
range polarization stripping in a loose intermediate where the
H2 and hydrocarbon moieties retain their identities.
To distinguish between the proton transfer and direct mech-
When the drift tube results for the N2H+/propene branching
ratios in Figure 2 are compared to the branching ratio for an
+
anism, we note in Table 3 that in all the reactions of D3 in
+
this work, except the reaction with propane, all the product ions
retained one D atom. The loss of H2 vs HD or CH4 vs CH3D
from the excited reaction intermediates produced by the initial
D+ transfer intermediates followed a statistical distribution
showing full hydrogen scrambling.
H3 reactant, it is apparent that the increased energy of the
reactant ions increases dissociation and narrows the gap between
+
the two product distributions. While the C2H3 product is not
seen in the N2H+ reaction at thermal energies, it constitutes
+
∼12% at -275 V, about half the value observed with the H3

Proton Transfer Reactions: Astrochemical Implications
J. Phys. Chem. A, Vol. 106, No. 42, 2002 9753
assume that radiative stabilization is not significant, at least for
the small ions with few vibrational modes. If stabilization of
(BH+)* is in fact negligible, then all the complexes with
sufficient energy to dissociate do dissociate. Conversely, the
observed nondissociated BH+ products are due to low-energy
(BH+)* intermediates that were formed with too little energy
to dissociate. These low-energy intermediates are formed by
the removal of the proton transfer exothermicity in inactive
modes, especially translation.35
For these reasons, the distribution of energy released by
proton transfer is of central interest. The internal energy of a
(BH+)* intermediate formed in reaction 1 can be calculated from
the energy balance of reaction 1 above:
Figure 2. Variation of branching ratio for the N2H+/propene reaction
Evib,rot(BH+)* + Etrans(BH+)* + Evib,rot(A) + Etrans(A) )
Ethermal (AH+) + Ethermal(B) + EPT (17)
as the voltage of the drift tube is altered. The line marked 43+ is for
+
+
the C3H7 ion, the line marked 41+ is for the C3H5 ion, and the line
+
marked 27+ is for the C2H3 ion. The results were obtained in 0.34
Torr of helium.
Here EPT is the exothermicity of the proton-transfer reaction at
o
+
0 K (i.e., ∆H0 ), and the other terms on the right-hand side are
ion. The C3H5 product is, as expected, less important at the
thermal energies of the reactants. The sum Etrans(BH+)* + Evib,rot
-
higher energies, while the C3H3+ product channel increases. This
indicates that the differences observed between the N2H+ and
(A) + Etrans(A) represents Einactive, the energy in modes that
(BH+)* cannot access for dissociation. For the present estimates,
we assume that, for small reactants with few vibrational modes
and for highly exothermic proton transfer, the thermal energies
of the reactants are negligible compared with EPT.
+
H3 proton-transfer products of C3H6 in Table 2 may be
eliminated if sufficient energy is placed in the N2H+ ion. The
N2H+/propyne system was also investigated by use of the drift
+
tube. As the voltage on the drift tube was increased, the C3H3
Neglecting the thermal energies Evib,rot + Etrans of the reactants
AH+ and B and Evib,rot(A), the energy in the active modes of
the intermediate ion is given by
ion was found to increase from ∼3% at thermal energies up to
12% at -250 V, which corresponds to 5 V cm-1 at 0.34 Torr.
Again this appears to show that energy in the N2H+ reactant
ion, possibly both as translation and vibration, can contribute
to dissociation. These observations are complementary to those
of Figure 1 and the results reported by Spanel et al.10 Both of
these sets of results show more dissociative channels as the
difference in PA between the proton acceptor and proton donor
increases.
(8) Energy Deposition and Product Distribution. In many
of the present reactions, and many other exothermic dissociative
reactions, a large fraction of the product ions do not decompose,
although the exothermicity can deposit more than enough energy
in the intermediate (BH+)* than is required for dissociation.
Evib,rot(BH+)* ) EPT - Einactive
(18)
Referring to Figure 3, the excess internal energy in the active
modes of Eex(BH+)* is the energy over Edissoc(BH+) + Ebarrier
,
where Edissoc(BH+) is the energy of dissociation of BH+ to form
the fragments F+ + N at 0 K. This excess energy is given by
subtracting Edissoc + Ebarrier from both sides of eq 18:
Eex(BH+)* ) Evib,rot(BH+)* - Edissoc(BH+) - Ebarrier
)
EPT - Einactive - Edissoc(BH+) - Ebarrier (19)
+
For example, 25-35% of the products of the reaction of H3
with cyclopropane remain as C3H5+, although dissociation to
Note also that the overall dissociative proton transfer (reac-
tions 1 + 2b) is equivalent thermochemically to proton transfer
to form BH+, followed by dissociation. Therefore the energy,
c-C3H3 is 86 kJ mol-1 exothermic. Similarly, in the H3+/2-
+
butene reaction, 35% of the ion products do not dissociate
although dissociative channels exist that are 252 and 226 kJ
mol-1 exothermic.
EDPT, of the overall dissociative proton-transfer reaction is given
by
The lack of dissociation may result from the release of a
fraction of the exothermicity of proton transfer in modes that
are not available for the dissociation of the (BH+)* intermediate.
Smith and Futrell examined this aspect of proton transfer by
comparing the reactions of KrH+, H3+, and O2H+, whose neutral
components have very similar proton affinities of 424.6, 422.3,
and 421.0 kJ mol-1, respectively, with the hydrocarbons CH4,
C2H4, and C2H6. In the most relaxed states of the reactant ions,
the product ratios for all ions are very similar. This indicates
that no significant energy is carried away in the vibrational
modes of H2 and O2, as krypton is an atomic species. The
exothermicity seems to be removed primarily as relative
translation between the product hydrogen molecule and car-
bonium ion.35
EDPT ) EPT - Edissoc(BH+)
(20)
(21)
Therefore from eqs 19 and 20
Eex(BH+)* ) EDPT - Einactive - Ebarrier
In the absence of a barrier, the excess energy of Eex(AB+)*
is equal simply to the exothermicity of dissociative proton-
transfer reaction minus the energy removed in the inactive
modes. According to the above considerations, (BH+)* dissoci-
ates if Eex(BH+)* > 0; i.e., if Einactive < EDPT - Ebarrier
.
Conversely, the reaction yields the undissociated proton-transfer
The distribution of energy may be estimated more quantita-
tively, along the lines applied by Fiaux et al.21 to the reaction
of H3+ with ethylene. The lack of pressure effects on the product
distributions suggests that collisional stabilization in reaction
2a is not significant. For the present considerations, we also
product BH+ if Eex(BH+)* < 0; i.e., if Einactive > EDPT - Ebarrier
.
If the barrier is negligible, the amount of energy removed in
the inactive modes to yield BH+ can be found simply from the
reaction thermochemistry, i.e., Einactive > EDPT. The fraction of
reactions where this amount of energy is removed can be found

9754 J. Phys. Chem. A, Vol. 106, No. 42, 2002
Milligan et al.
Figure 3. Schematic energy diagram for dissociative proton-transfer reactions AH+ + B fm (BH+)* fm F+ + N. The energy terms are defined in
the text.
from the ratio [BH+]/([BH+] + [F+]) when the product
distributions are known.
calculations can be useful when there are competitive dissocia-
tion channels whose ratio is sensitive to the energy. Examples
are the reactions in Table 2 where several dissociation channels
exist and the product distribution changes in going from N2H+
to H3+ reactions. In such cases RRKM calculations can match
the observed product distributions to the energy distribution in
the (BH+)* population and from this can identify the distribution
of EPT in the active and inactive modes.
For example, in the reaction of N2H+ with CH3CHdCH2,
the dissociative reaction is exothermic by 110 kJ mol-1 but 55%
of the (C3H7+)* ions do not dissociate. Therefore, in 55% of
the reactions, more than 110 kJ mol-1, i.e., 42% of the 258 kJ
mol-1 proton-transfer exothermicity, is removed in the inactive
modes. Similar estimates may be made for other reactions in
Table 2 where both dissociative and nondissociative channels
exist. In the reactions of N2H+ with c-C3H6, >49% of the
exothermicity is removed in 10% of the reactions; with 2-butene,
>57% is removed in 10% of the reactions.
Conclusions
Exothermic proton-transfer reactions from H3O+, N2H+, and
H3+ to hydrocarbons are dissociative to various extents, produc-
ing stable protonated ions, BH+, and fragments F+. Common
dissociation pathways following exothermic proton transfer are
hydrogen loss, methane loss, and the loss of other small alkanes
or alkenes, with the dissociative channels increasing with
increasing exothermicity of proton transfer. The main question
in such reactions is the distribution of the reaction exothermicity
in the reaction intermediates, which control the product distribu-
tions.
Similar considerations can be applied to reactions where only
nondissociative reactions are observed although exothermic
dissociation channels exist. However, they may overestimate
the removed energy. For example, in the reaction of N2H+ with
toluene, which is 100% nondissociative, >82% of the exother-
micity would need to be removed in all the reactions to prevent
H2 loss. This seems unlikely and suggests that the lack of
dissociation is due to an energy barrier.
Deuterium labeling in this and preceding work10,21,32-35 shows
that most reactions in this paper are best characterized by proton
transfer to form excited (BH+)* intermediates. A fraction of
these intermediates dissociates at a rate that allows hydrogen
scrambling. A lack of pressure effects on product distributions
suggest that the decomposing fraction of (BH+)* is not stabilized
by collision and dissociates faster than the collision time in the
SIFT (∼3 × 10-8 s). The lack of collisional stabilization
indicates that the intermediates (BH+)* that do not dissociate
within the time frame of the SIFT experiment contain too little
energy to do so. On this basis the energy distribution upon
proton transfer can be estimated, showing that up to 40-60%
of the reaction exothermicity may be removed in several
reactions (less if barriers exist) as translational energy, in
agreement with the conclusions of previous workers.21,32-35 In
reactions where no stable protonated ion exists, as is the case
for propane, unimolecular decay occurs instantaneously without
allowing for hydrogen scrambling. RRKM calculations on
reactions with competing channels may define the energy
distributions more quantitatively.
Where barriers to dissociation exist, the excess energy
required for dissociation reflects the difference between ground-
state BH+ and the energy required for its dissociation including
the barrier (see Figure 3). Clearly, if a barrier exists, then less
energy needs to be removed in the inactive modes to prevent
dissociation. Therefore, the examples give a lower limit to the
energy retained in (BH+)* and an upper limit to the energy
removed from the dissociating population. For example, the
+
+
dissociation of C3H7 to C2H5 + H2 was found by several
investigators to involve a barrier (reverse activation energy)36-38
of 70 kJ mol-1 39 Therefore, only 40 kJ mol-1, rather than 110
.
kJ mol-1, needs to be removed in 55% of the reactions of N2H+
+
+ CH3CHCH2 to obtain the C3H7 ion.
The other approximation above was neglecting the energy
of (BH+)* resulting from the thermal energies of the reactants.
This correction has the reverse effect on the considerations of
energy distribution and increases the energy that must be
removed in the inactive modes to prevent dissociation. This
correction can be significant: for example, a significant fraction
of the C3H7+ population at 300 K has vibrational energy up to
20 kJ mol-1 39
.
For smaller hydrocarbons including alkanes, alkenes, alkynes,
and cyclic species, dissociative reactions are usually observed
if the exothermicity of a particular channel exceeds 100 kJ
mol-1. In some cases, more highly exothermic channels are less
prominent unless there is no other available exothermic reaction
pathway.
A more quantitative calculation of the energy deposited in
(BH+)* may be obtained if the potential energy surfaces are
known. If the potential energy surfaces and transition-state
parameters are known, RRKM calculations can relate the energy
content of (BH+)* to the observed product ratios. These

Proton Transfer Reactions: Astrochemical Implications
J. Phys. Chem. A, Vol. 106, No. 42, 2002 9755
(7) Aquilanti, V.; Volpi, G. G. J. Chem. Phys. 1966, 44, 2306 and
3574.
In aromatic hydrocarbons, as the molecules increase in size,
the tendency for dissociative proton transfer decreases as the
energy in (BH+)* is distributed in many vibrational modes. The
aromatic hydrocarbons observed, such as benzene and the
polycyclic aromatics naphthalene and methylnaphthalenes,
exhibit no dissociation channels. Notably, while the reaction
(8) Burt, J. A.; Dunn, J. L.; McEwan, M. J.; Sutton, M. M.; Roche, A.
E.; Schiff, H. I. J. Chem. Phys. 1970, 52, 6062.
(9) Lee, H. S.; Drucker, M.; Adams, N. G. Int. J. Mass Spectrom. Ion
Proc. 1992, 117, 101.
(10) Spanel, P.; Smith, D.; Henchman, M. Int. J. Mass Spectrom. Ion
Proc. 1995, 141, 117.
(11) Smith, D.; Spanel, P.; Millar, T. J. Mon. Not. R. Astron. Soc. 1994,
266, 31.
(12) Spanel, P.; Pavlik, M.; Smith, D. Int. J. Mass Spectrom. Ion Proc.
1995, 145, 177. (b) Spanel, P.; Smith, D. Int. J. Mass Spectrom. 1998,
181, 1.
(13) Smith, D.; Spanel, P.; Mayhew, C. A. Int. J. Mass Spectrom. Ion
Proc. 1992, 117, 457.
(14) Milligan, D. B.; Fairley, D. A.; Freeman, C. G.; McEwan, M. J.
Int. J. Mass Spectrom. 2000, 202, 351.
(15) Smith, D.; Spanel, P. Int. J. Mass Spectrom. Ion Proc. 1993, 129,
163.
(16) Anicich, V. G. J. Phys. Chem. Ref. Data 1993, 22, 1469.
(17) Su, T.; Chesnavich, W. J. J. Chem. Phys. 1982, 72, 5183.
(18) Mackay, G. I.; Tanaka, K.; Bohme, D. K. Int. J. Mass Spectrom.
Ion Phys. 1977, 24, 125.
+
of H3 with toluene is partly dissociative, the reaction with
methylnaphthalene is not, showing the stabilizing effect of the
larger molecular size. The lack of dissociation may be due to
radiative stabilization of (methylnaphthaleneH+)* or alterna-
tively, its longer lifetime allowing collisional stabilization. Both
of these stabilization processes are facilitated by the large
number of active modes. We also note that the C7H7+ ion formed
by H2 loss from toluene may have the tropylium structure, but
in methylnaphthalene a tropylium ion would be fused to a
benzene ring, which may be destabilizing. In any event, the
lack of dissociation of methylnaphthalene suggests that H3+ may
not be able to dissociate alkylated or unalkylated PAHs.
The interstellar medium and the atmospheres of Jovian planets
are dominated by hydrogen, and ionizing radiation rapidly
(19) Kim, J. K.; Theard, L. P.; Huntress, W. T. Int. J. Mass Spectrom.
Ion Phys. 1974, 15, 223.
(20) McIntosh, B. J.; Adams, N. G.; Smith, D. Chem. Phys. Lett. 1988,
148, 143.
+
produces H3 that is likely to be the major ionic reagent. The
+
dissociative reactions of H3 with nonaromatic hydrocarbons
(21) Fiaux, A. S.; Smith, D. L.; Futrell, J. H. J. Am. Chem Soc. 1976,
98, 5773.
(22) Mackay, G. I.; Tanner, S. D.; Hopkinson, A. C.; Bohme, D. K.
Can. J. Chem., 1979, 57, 1518.
(23) Mallard, W. G., Linstrom, P. J., Eds. NIST Chemistry WebBook;
NIST Standard Reference Database Number 69, February 2000; National
Institute of Standards and Technology: Gaithersburg, MD 20899 (http://
webbook.nist.gov).
(24) Spanel, P.; Smith, D. Int. J. Mass Spectrom. Ion Proc. 1998, 181,
1.
(25) Bohme, D. K. In Interactions between Ions and Molecules; Ausloos,
P., Ed.; NATO Advanced Study Institute: Biarritz, France, 1974; p 489.
(26) McEwan, M. J.; McConnell, C. L.; Freeman, C. G.; Anicich, V.
G. J. Phys. Chem. 98, 5068.
(27) Fairley, D. A.; Milligan, D. B.; Wheadon, L. M.; Freeman, C. G.;
Maclagan, R. G. A. R.; McEwan, M. J. Int. J. Mass Spectrom. Ion Proc.
1999, 185-7, 253.
can be a major mechanism for converting more-saturated to less-
saturated species by H2 loss and for converting larger to smaller
molecules by the loss of CH4 or other small hydrocarbons. On
the other hand, aromatics are observed to be stable against
decomposition by H3+ and also by other protonating ions where
proton transfer is less exothermic. We also note that these
processes are paralleled by electron impact ionization, where
the unsaturated hydrocarbons fragment more readily while PAHs
yield significant amounts of molecular ions. Ionizing reactions,
including dissociative ion-molecule reactions of H3+, therefore
favor the accumulation of unsaturated and aromatic molecules
in ionizing and hydrogen-dominated astrochemical environ-
ments.
(28) Ausloos, P.; Lias, S. G.; Buckley, T. J.; Rogers, E. E. Int. J. Mass
Spectrom. Ion Proc. 1989, 92, 65.
(29) Giles, K.; Adams, N. G.; Smith, D. Int. J. Mass Spectrom. Ion Proc.
1989, 89, 303.
Acknowledgment. We thank the Marsden Fund for financial
support of this work. P.F.W. thanks the University of Canterbury
for the award of a postdoctoral fellowship.
(30) Scott, G. B. I.; Fairley, D. A.; Freeman, C. G.; McEwan, M. J. J.
Phys. Chem. A. 1997, 101, 4973.
(31) Gardner, F. F.; Winnewisser, G. Astrophys. Lett. 1995, 195, L127.
(32) Fiaux, A.; Smith, D. L.; Futrell, J. H. Int. J. Mass Spectrom. Ion
Phys. 1974, 15, 9.
(33) Smith, D. L.; Futrell, J. H. J. Phys. B 1975, 8, 803.
(34) Smith, R. D.; Futrell, J. H. Int. J. Mass Spectrom. Ion Phys. 1976,
20, 71.
(35) Smith, R. D.; Futrell, J. H. Int. J. Mass Spectrom. Ion Phys. 1976,
20, 347.
(36) Holmes, J. L.; Osborne, A. D. Org. Mass Spectrom. 1978, 13, 133.
(37) Trager, J. Int. J. Mass Spectrom. Ion Phys. 1980, 32, 309.
(38) Almlof, J.; Hvistendal, G.; Uggerud, E. Chem. Phys. 1984, 90, 55.
(39) Boyd, R. K. Int. J. Mass Spectrom. Ion Phys. 1984, 55, 55.
References and Notes
(1) Herbst, E. In AdVances in Gas-Phase Ion Chemistry; Adams, N.
G., Babcock, L. M., Eds.; JAI Press Inc.: Greenwich, CT, 1998; Vol. 3, p
1.
(2) Gaballe, T. R.; Oka, T. Nature 1996, 384, 334.
(3) Drossart, P.; Maillard, J. P.; Caldwell, J.; Kim, S. J.; Watson, J. K.
G.; Majewski, W. A.; Tennyson, J.; Miller, S.; Atreya, S. K.; Clarke, J. T.;
Waite, J. H.; Wagener, R. Nature 1989, 340, 539.
(4) Hunter, E. P. L.; Lias, S. G. J. Phys. Chem. Ref. Data 1998, 27,
413.
(5) Wexler, S. J. Am. Chem. Soc. 1963, 85, 272.
(6) Munson, M. S. B.; Franklin, J. L.; Field, F. H. J. Am. Chem. Soc.
1963, 85, 3584.