H3++O: An experimental study

Article abstract of DOI:10.1016/S0009-2614(00)00154-8

We have measured the reaction rate coefficient and branching ratios for the H₃⁺+O atom reaction using a flowing afterglow/selected ion flow tube operating at room temperature (295±5 K). The measured rate coefficient was k=(1.2±0.5)×10^-9 cm³ s^-1 and the branching ratios were OH⁺+H₂, (70%) and H₂O⁺+H (30%). This reaction has some relevance to the synthesis of water in interstellar clouds.

Full text of DOI:10.1016/S0009-2614(00)00154-8

2
4 March 2000
Chemical Physics Letters 319 Ž2000. 482–485
www.elsevier.nlrlocatercplett
q
H qO: an experimental study
3
Daniel B. Milligan, Murray J. McEwan )
Department of Chemistry, UniÕersity of Canterbury, Christchurch, New Zealand
Received 15 October 1999; in final form 10 January 2000
Abstract
q
We have measured the reaction rate coefficient and branching ratios for the H qO atom reaction using a flowing
3
afterglowrselected ion flow tube operating at room temperature Ž295"5 K.. The measured rate coefficient was
y
9
3
y1
q
q
ksŽ1.2"0.5.=10 cm s and the branching ratios were OH qH , Ž70%. and H O qH Ž30%.. This reaction has
2
2
some relevance to the synthesis of water in interstellar clouds. q 2000 Elsevier Science B.V. All rights reserved.
1
. Introduction
The reaction between H3 and atomic oxygen is
mic binary channels available plus the possibility of
a termolecular association channel:
q
q
q
y1
H qO ™ OH qH q63.6 kJ mol
,
Ž
Ž
1a
1b
.
.
3
2
relevant to interstellar cloud chemistry. Model abun-
dances of atomic oxygen in dense clouds has been
estimated to be as high as 2=10 relative to H at
q
y1
™
H O qHq163.9 kJ mol
,
y4
2
2
y4
early cloud times w1x and 1.5=10
relative to H2
M
™
q
y1
H O q760 kJ mol
.
Ž
1c
.
at steady-state cloud conditions at 10 K w2x. Further,
3
q
the strategic H3 ion has been identified recently in
We report here the results of a flowing after-
dense clouds in the interstellar medium w3x. The
glowrselected ion flow tube ŽFArSIFT. investiga-
tion of reaction Ž1. in which we have measured the
rate coefficient and determined the products of reac-
tion.
reaction between Hq3 and atomic oxygen provides a
pathway for synthesising water in interstellar clouds.
The reaction is, however, difficult to study experi-
mentally with only one previous measurement hav-
ing been reported by Fehsenfeld w4x. In this study, a
fast reaction was reported having a rate coefficient
ksŽ8.0"4.=10
action were not determined. There are two exother-
y10
3
y1
cm s . The products of re-
2
. Experimental
The FArSIFT apparatus used in these studies and
operating at room temperature Ž295"5 K., has been
described in a recent paper w5x. The flowing after-
glow FA source section of this equipment is an
essential feature of this study and will be mentioned
)
Ž
.
Corresponding author. Fax: q64-3-3642-110; e-mail:
m.mcewan@chem.canterbury.ac.nz
0
009-2614r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved.
PII: S0009-2614Ž00.00154-8

D.B. Milligan, M.J. McEwanrChemical Physics Letters 319 (2000) 482–485
483
briefly here. In the one previous study of reaction
q
Ž1., H3 ions were formed in the flow tube by the
3
reaction of He Ž2 S. metastables with H w4x.
2
3
q
He
Ž
2 S
.
qH ™ H qHeqe ,
Ž
Ž
2
3
.
.
2
2
q
q
H qH ™ H qH .
2
2
3
It was the presence of H2 in the flowing afterglow
that prevented identification of the products as both
q
q
possible products, OH and H O are converted
2
q
rapidly to H O in the presence of H .
3
2
q
In this work, we formed the H3 with a mi-
crowave discharge on a hydrogen carrier gas at the
upstream end of the FA source reaction tube operat-
ing at 0.70 Torr. A small fraction of the ions formed
are sampled through a 2.0 mm orifice in a nose cone
at the downstream end of this flow tube w5x. These
q
Fig. 1. The ions observed in the reaction tube after injecting H3
formed from a microwave discharge in H2 in the flowing after-
q
glow source. Only small amounts of HeH are present. The
q
q
signal at 19 is H O and 29 is N H arising from trace
3
2
ions were directed, via an array of electrostatic lenses
and focussed into the entrance aperture of a
impurities of H2O and N2 , respectively.
y6
O atoms were formed by subjecting a 5% mixture
quadrupole mass filter operating at ;1=10 Torr.
q
of O in helium to a microwave discharge: a method
The H3 ions were mass selected and then transferred
2
that has been successfully used previously w9x. A
through a second lens array and focussed into the
entrance aperture of a Venturi nozzle for transmis-
glass wool plug was added downstream of the dis-
charge to remove metastable species from the gas
stream w10x. The technique more commonly used for
q
sion into the SIFT reaction tube w5x. The H3 ion
swarm was transported along the SIFT reaction tube
in a helium carrier gas. By using the FA source in
O atom production in flow tubes is the titration
q
w
x
reaction between N atoms and NO 10,11 . Although
it has the advantage of providing an absolute mea-
sure of O atom densities, it could not be used in the
present study because the two principal product ions
this way, we were able to create a H3 ion swarm in
a helium carrier gas, free from substantial amounts
of vibrational energy, that has been the traditional
q
bane of H3 studies w6,7x. One reaction providing
q
q
q
of reaction 1 , OH and H O , undergo further
Ž .
evidence for vibrational excitation in H3 is the
2
reaction with N, N and NO, thus obscuring product
reaction with argon for which only those levels
above the
2
ion identification.
q
q
q
y1
The reaction between CH
3
and O was used in
H
Ž
n
.
qAr ™ ArH qH y55 kJ mol
4
Ž .
3
2
this work to calibrate the O atom densities. The
reaction rate coefficient for this reaction is well
three lowest vibrational states Žthe ground state and
Õs1 for the bending and breathing modes. have
y10
3
y1
established as ks4.4=10
cm s w4,9x and we
y1
sufficient energy to meet the 55 kJ mol deficiency
have also previously measured the rate coefficient
q
from ground vibrational state H w8x. Less than 10%
3
using the NqNO titration reaction as ks4.1=
q
of the injected H3 ion swarm showed any reactivity
with Ar. The other difficulty caused by injecting H3
y10
3
y1
10
cm s w12x. As with most reactions of atoms
q
generated from discharges, the errors inherent in
monitoring the atom flux give an uncertainty in the
rate coefficient of "40%. The branching ratios are
considered accurate to "30%.
into a stream of helium is that the ion energies
required for injection can be sufficient to generate
q
HeH during the injection process via the endoergic
reaction
q
q
y1
H qHe ™ HeH qH y245 kJ mol
.
5
Ž .
3
2
3
. Results
By careful control of the injection energies, we were
q
able to drift H3 ions into a helium carrier gas free
The product distributions for reaction Ž1. were
q
from significant HeH formation ŽFig. 1..
found in the usual way by plotting the fraction of

4
84
D.B. Milligan, M.J. McEwanrChemical Physics Letters 319 (2000) 482–485
q
thought earlier that the reaction between H qN
3
q
underwent a similar H transfer w14x but have since
2
discovered no reaction occurs w5,12x.
The experimental results obtained here provide
some confirmation of the success of ion–molecule
dynamic calculations using ab initio potential energy
surfaces. Bettens et al. w15x have recently constructed
q
a potential energy surface for triplet state OH3 at the
MP2r6-311G Ž2d, p. level of theory. A number of
classical trajectories were performed with the sur-
face, from which they calculated an overall rate
q
Fig. 2. The product distribution graph for the reaction of H qO
3
atoms. The data have been corrected for the residual background
q
y10
3
signal of H O .
coefficient for reaction Ž6. of ks9.4=10
cm
2
y1
q
s
at 300 K and a branching ratio of 94% ŽOH q
q
H . and 6% ŽH O qH.. The overall agreement
2
2
each product ion as a function of O atom flow Žsee
Fig. 2 for a typical product distribution determina-
tion.. A very small signal of H O from photo
ionization was always present. This signal did not
change with the reactant HerO flow and was sub-
tracted from the H O product channel. The mea-
between theory and experiment is thus remarkably
good. Bettens et al. also comment in their theoretical
q
w x
study 15 that they may have underestimated the
2
q
branching ratio to the H O qH channel Žreaction
2
Ž6b.. because of surface hopping to the singlet sur-
face. Our results indicate that this may indeed be a
factor even when the 30% uncertainty in the mea-
sured branching ratio is taken into account.
2
q
2
sured product distributions were
0
.70
q
q
H qO ™ OH qH ,
Ž
Ž
6a
6b
.
.
3
2
0
™
.30
q
H O qH ,
2
5
. Conclusions
and the measured rate coefficient was ksŽ1.2"
y9
3
y1
0
.48.=10 cm s which is in reasonable agree-
ment with the one earlier measurement of ksŽ8"
We have been able to measure the product ratio of
q
the important interstellar cloud reaction, H qO for
3
y10
3
y1
4
.=10
cm s w4x. No evidence for any associ-
the first time. A comparison of the observed results
with the results predicted using potential energy
surfaces generated by ab initio methods is a confir-
mation of the predictive capability of such tech-
niques in systems with small numbers of atoms.
ation channel was found.
In the course of this study we also remeasured the
q
y10
rate coefficient between H qO as ks5.9=10
3
2
3
y1
cm s
q
q
H qO ™ HO qH ,
7
Ž .
3
2
2
2
which agrees well with several previous measure-
ments w13x.
Acknowledgements
We thank the Marsden Fund for financial support
of this work.
4
. Discussion
The observation of the 30% channel to H O qH
q
2
References
Žreaction Ž6b.. is the first experimental evidence for
a reaction of H3 where H2 is transferred rather than
proton transfer or charge transfer. Reaction Ž6. is the
fastest ion–O atom reaction we have observed to
date w11,12x, proceeding at, or close to, the Langevin
collision rate Žk s1.3=10
q
q
w1x T.J. Millar, P.R.A. Farquhar, K. Willacy, Astron. Astrophys.
Suppl. Ser. 121 Ž1997. 139.
w2x H.-H. Lee, R.P.A. Bettens, E. Herbst, Astron. Astrophys.
Suppl. Ser. 119 1996 111.
Ž
.
y9
3
y1
cm s .. We had
w3x T.R. Gaballe, T. Oka, Nature ŽLondon. 384 Ž1996. 334.
L

D.B. Milligan, M.J. McEwanrChemical Physics Letters 319 (2000) 482–485
485
w4x F.C. Fehsenfeld, Astrophys. J. 209 Ž1976. 638.
w5x D.B. Milligan, D.A. Fairley, C.G. Freeman, M.J. McEwan,
Int. J. Mass Spectrom. Ž1999., submitted.
w11x G.B.I. Scott, D.A. Fairley, D.B. Milligan, C.G. Freeman,
M.J. McEwan, J. Phys. Chem. A 103 Ž1999. 7470.
w12x G.B.I. Scott, D.B. Milligan, D.A. Fairley, C.G. Freeman,
M.J. McEwan, J. Chem. Phys. Ž2000., in press.
w13x V.G. Anicich, J. Phys. Chem. Ref. Data 22 Ž1993. 1469.
w14x G.B.I. Scott, D.A. Fairley, C.G. Freeman, M.J. McEwan,
Chem. Phys. Lett. 269 Ž1997. 88.
w6x W.T. Huntress, M.T. Bowers, Int. J. Mass Spectrom. Ion
Phys. 12 Ž1973. 1.
w7x D. Smith, P. Spanel, Chem. Phys. Lett. 211 Ž1993. 454.
w8x T. Gougousi, R. Johnsen, M.F. Golde, Int. J. Mass Spectrom.
Ion Processes 149,150 Ž1995. 131.
w15x R.P.A. Bettens, T.A. Hansen, M.A. Collins, J. Chem. Phys.
w9x A.A. Viggiano, R.A. Morris, J.F. Paulson, E.E. Ferguson,
Ž1999., in press.
Chem. Phys. Lett. 148 Ž1988. 296.
w10x P.D. Goldan, A.L. Schmeltekopf, F.C. Fehsenfeld, H.I. Schiff,
E.E. Ferguson, J. Chem. Phys. 44 Ž1966. 4095.