International Journal of Mass Spectrometry and Ion Processes, 98 (1990) 179-190 Elsevier Science Publishers B.V., Amsterdam - Printed in The Netherlands

179

VARIABLE-TEMPERATURE SELECTED ION FLOW TUBE STUDIES OF THE REACTIONS OF Ar+‘, Arz’ AND ArH,f (n = 1-3) IONS WITH H,, HD AND D, AT 300 K AND 80 K

D.K. BEDFORD and D. SMITH

School of Physics and Space Research, University of Birmingham, P.O. Box 363, Birmingham B15 2TT (U.K.)

(Received 4 February 1990)

ABSTRACT

The reactions of Ar+’ , Art’, ArH+ (ArD+ ), ArH: (ArHD+ , ArD: ) and ArH: with H,, HD and Dz have been studied at 300 K and 80 K using our variable-temperature selected ion flow tube (VT SIFT). The Ar+’ reactions proceed via parallel H-atom abstraction (major channel) and charge transfer, and the rate coefficients, k, are about 0.55 times the collisional rate coefficients, k,, at both temperatures for the three reactions. The ArH+(ArD+) reactions proceed via proton (deuteron) transfer, and the k values are about 0.6-k, at 300 K and about 0.75-k, at 80 K. The Ar:’ reactions proceed via two parallel channels, for example, the Hz reaction produces ArH+ (atom abstraction) and ArH” (Ar atom/Hz molecule switching); the analogous reactions occur with HD and D,. Whilst the atom abstraction channel is favoured at 300K (in the ratio of about 70:30) the two channels are comparable at 80K for all three reactions. The overall k value for these reactions is about 0.3 -k, at 300 K increasing to about (0.5-0.6)-k, at 80K; in effect, the switching reactions are significantly enhanced at the low temperature. The ArH: reaction with Hz produces ArH: ; similarly ArH, DC, ArHD: and ArDT are produced in the other reactions. The k values are about 0.7-k, at both temperatures. From these studies, the dissociation energy of Ar*H$ (to Ar + Ht’) is shown to be > 1.04eV and that of ArH: (to Ar + H:) is shown to be within the range 0.29-0.36eV.

INTRODUCTION

A recent detailed study in our laboratory of the reactions of the rare gas molecular ions Kr2+. and Xe,+. [ 1,2] revealed the interesting phenomenon that the ions undergo “switching reactions” with many molecules, M, producing the ions KrM+’ and XeM+‘; however, if the ionization energy of M exceeds the recombination energy of Kr:’ and Xel’ , then the process of charge transfer cannot also occur. A subsequent study of the reactions of the KrM+’ and XeM+’ ions with molecules M [3] revealed the extremely interesting results that the dimer ions M+ *M were efficiently formed (ions such as the water dimer ion, M+ *H,O, and the methane dimer ion CH:’ -CH,).

0168-l 176/90/$03.50 0 1990 Elsevier Science Publishers B.V.

180

The recombination energy of Ar,+’ (RE(Ar: ‘) = 14.4 eV) exceeds the ionization energies of most molecules and so a common reaction mechanism for this ion is likely to be charge transfer. However this process cannot occur between Ar,f ’ and H, and N2 (by virtue of the high ionization energies of these molecuies) and so “switching” becomes a possible reaction process. We have therefore commenced a study of the reactions of Arc’ in our laboratory, initially studying the reactions with H,, HD and D, at 300 K and 80 K in our variable-temperature selected ion flow tube (VT SIFT) apparatus. Some previous studies of the reactions of Ar: ’ ions with H, have been carried out [4,5,6], the most recent being by Shul et al. [7] in a SIFT apparatus at 300 K, who placed a limit on the proton affinity of Ar atoms from their data relating to the Ar: ’ and H, reaction. During the course of the present studies we discovered a route for the efficient production of the remarkable ionic species ArH: , which is known to possess a very large permanent dipole moment [8]. In parallel with these studies, we have also studied the reactions of Ar+ ions with H,, HD and D,, also at 300K and 80 K, to clarify the somewhat conflicting data in the literature relating to these reactions.

EXPERIMENTAL DETAILS

The measurements discussed here were made using the VT SIFT technique, which has been described in detail elsewhere [9]. For most of the reactions in this study, the reactant ions Ar+’ and Ar: ’ were produced in a high pressure ion source containing pure argon, the pressure of which was varied to optimise the production of either the atomic or the molecular ion. To produce ArH+ or ArD+ , H, or D, were mixed with Ar in the source. (The ArH;’ ions-and the deuterated and partially deuterated analogues-could not be generated in the source in sufficient concentration and so they were made in the flow tube by the injection of Ar: ‘, which then reacted with H,, (HD, D2) to produce the desired ions.) Primary ions from the source were selected according to their mass-to-charge ratio by a quadrupole mass spectrometer, before injec- tion into the helium carrier gas which passes along the flow tube at a pressure of about 0.5 Torr. Entry ports along the flow tube permit the introduction of reactant neutral species into the flowing ion/carrier gas mixture. The primary reactant ions and the product ions are sampled via a downstream orifice and, after selection by another quadrupole mass filter, are counted for a range of values of neutral reactant flow. The flow tube is temperature controlled within a vacuum jacket; the present measurements were made at room temperature (300 K) and at 80 K using liquid nitrogen to cool the flow tube.

The rate coefficients for reactions of ions which can be directly injected into the flow tube from the ion source are calculated from the rate of decrease of the primary ion count-rate as a function of the flow rates of reactant neutral

181

species. Branching ratios are determined from the relative count-rates of the product ions after correcting, if necessary, for mass discrimination by the quadruople mass filter. For the cases where the reactant ion cannot be efficiently formed in the ion source and injected directly into the flow tube (e.g. ArHD+‘), the rate coefficient of the secondary reaction (i.e. ArHD+’ + HD in this case) is obtained from the variation of the ion count rate (ArHD+) as it increases at low neutral flow rates during its formation from the primary reaction (Ar: ’ + HD), and then decreases as it reacts away as the neutral flow (in this example HD) is increased further. An algorithm is used which iteratively computes the reactant ion count-rate/neutral flow-rate curve, which requires the input of the rate coefficient for the primary reaction (Ar2+- + HD) which produces the ion (ArHD+‘) and allows the required rate coefficient for the secondary reaction to vary until a satisfactory tit to the experimental data points is achieved. Good agreement was found between rate coefficients determined by this method for ArH+(ArD+) + H2(D2) and those determined by the direct method (i.e. by direct injection into the flow tube of ArH+ (ArD+) as the primary ion), which gives credence to the rate coefficients computed by this method.

The effect of spin-orbit splitting of the ground electronic state of the Ar+’

ion (‘P1j2 y ‘G2> may well account for the quite large range in previously reported values of the rate coefficient for the Ar+’ + H, reaction. Separa- tion of the two states of Ar+ has not been achieved by chemical filtering (as has been done for the spin-orbit states of Kr+’ and Xe+’ [1,2]) because the energy difference between the spin-orbit states, 0.178 eV, is too small (cf. Kr+ and Xe+ with energy differences of 0.666 and 1.306 eV). In the SIFT measure- ments reported here, the evidence (given below) is that only ions in the lower energy state (2PJ,2) are present in the reaction zone of the SIFT. The absence of curvature in the Ar+’ decay curves obtained in the present experiments indicates that only one reacting species is present, unless both species have equal rate coefficients, k, for all the reactions studied, which is improbable when the k values are less than the collisional rate coefficients, k,. Superelastic collisions with electrons in the source is the most probable relaxation process for ions in the 2P,,2 state [IO]. Quenching in the source by collisions with Ar atoms is unlikely, since Liu and Conway [ 1 I] and Liao et al. [ 121 have shown the rate coefficient to be very small or zero for that process. Any Ar+’ (2P,,2) ions that are injected into the flow may be quenched in collisions with the helium carrier gas atoms, or in collisions with the reactant gases Hz, HD and D,-a process which both Hamdan et al. [13] and Rakshit and Warneck [14] have shown to be quite rapid (k,, for quenching of Ar+ (‘PIi2) ions, is l-2 x 10-gcm3s-‘). Our earlier work on Ar+’ + N, has also shown no evidence that Ar+’ (‘P,,,) is present in the flow tube [15].

182

RESULTS

Reactions of Ar+’ with H2, HD and D,

These reactions proceed by parallel atom abstraction (reaction la) and charge exchange (reaction lb) channels, e.g.

Ar+’ + H, + ArH+ + H’ (la)

+ Hz’ + Ar (lb)

The analogous reactions proceed with D,; with HD, both ArH+ and ArD+ are produced with equal probability in the atom abstraction reaction (see Table 1). The atom abstraction product ArH+ has been observed by all previous workers who have studied this reaction, but the charge exchange product H: has been detected less often. Ryan and Graham [16] observed 30% of H:‘, Smith et al. [17] 4% of H:‘, and Tanaka et al. [18] obtained cross-sections for reaction 1 using their TESICO technique which yielded 5% of H:’ (with a ratio between the cross-sections of 7: 1 for production by the 2P,,2:2Pj,2 states). Tanaka et al. [18] also investigated the reaction with D,, where the charge exchange product D:’ was only N 2% of the total product ions, independent of the spin-orbit state of the Ar+’ ion. Most recently, rate coefficients at 20K, 30K and 70K have been reported for Ar+ + H, by Rebrion et al. [19], using the CREWS technique. Combining their data with results at higher temperatures from other workers, they find a weak temperature-dependence for the overall rate coefficient k, for the H, reaction 1.

Unlike the TESICO experiments, the present measurements were made without state selection of the Ar+ ion although, as we have stated already, the evidence is that only the lower state Ar+’ (*P3,*) ions are present in the reaction zone of the flow tube. For the reactions of Ar+ ’ with H,, HD and D,, the charge exchange channels were clearly present, as well as the well-established atom abstraction channels. Table 1 gives the observed product ratios, the overall rate coefficients, k, = (k, + kb), the calculated Langevin (collisional) rate coefficients, k,, and the reaction exothermicities. It may be noted in all cases that k, I: 0.55-k,, that is k, is less than collisional in every reaction at both 300 K and 80 K, and also that k, scales with p-‘j2, where ~1 is the reduced mass of the reactants [20], indicating that there is no kinetic isotopic effect in these reactions. Our value of k, = 7.2 x lo-“cm3 s-’ for the reaction of Ar+’ with H, at 80 K is in excellent agreement with that of Rebrion et al. [ 191 at 70 K who found that k, = 7.1 x lo-“cm3 s-‘. Our value of k, = 8.6 x lo-” cm3 s-’ at 300 K is in good agreement with recent values obtained at 300K by Shul et al. [7], 8.2 x 10-‘“cm3s-‘, Hamdan et al. [13], k, = 8.8 x 10-‘“cm3s-‘, and Dotan and Lindinger [21], k, = 9.5 x 10-‘“cm3s-‘, but

183

TABLE 1

The measured total rate coeffkients, k, the Langevin (collisional) rate coeffkients, kc, and the ion-product ratios for the reactions of Ar+ with Hz, HD and D, at 300 K and 80 K. Also given are the exothermicities of the reactions, AE

Reaction cm3 s-l)

s-‘) 300 K 80K

Ar+ + H, --f ArH+ + H 85 93 + 1.33 + H: + Ar 15 7 8.6 7.2 15.2 f0.33

Ar+ + HD -+ ArH+ + D 44 45 -+ ArD+ + H 44 43 -

-+ HD+ + Ar 12 11 7.4 7.1 12.5 + 0.30

Ar+ + D, + ArDf + D 88 85 -

-+ D: + Ar 12 1.5 6.3 6.5 10.9 + 0.29

is somewhat greater than the values obtained by Adams et al. [22], k, = 7.4 x 10-‘“cm3s-‘, and Rakshit and Warneck [14], k, = 5.4 x lo-“cm3 SC’. Although our data indicate a weak but significant temperature dependence for k, for the Ar+’ + H, reaction (errors are + 10% on k), in excellent agreement with the work of Rebrion et al. [19], no significant temperature dependence is seen for the analogous reactions with HD and D, .

It is clear from the product ratios (Table 1) that there is neither a strong temperature effect on the branching ratios in all these reactions, nor any significant kinetic isotope effect at either 300 K or 80 K. The atom abstraction channel is dominant at both temperatures, with about 90% of the products going into that channel regardless of the degree of deuteration of the reactant neutral molecule. Further, as already stated, in the reaction with HD there is no bias towards either atom abstraction product, ArH+ or ArD+ . The atom abstraction channel is quite exothermic (2 1.33 eV, taking the proton affinity of Ar atoms, PA(Ar), to be > 3.69 eV [7]), the charge exchange channel less so (0.33 eV for the H, reaction lb). There is some suggestion of a systematic isotope effect in the Hl’ , HD+ ‘, D: ’ products at 80 K, with yields of 7, 11, and 15%, respectively. This apparent trend (which is not evident at 300K) is consistent with the increased phase space available to the products as the vibrational energy levels become more closely spaced for the heavier product ions.

Reactions of AI-H+ (ArD+) with H2 (HD, D,)

Both primary products, ArH+ and H2f, formed in reactions la and lb, undergo binary reactions with H2 to form the terminating ion Hz ; similarly

184

TABLE 2

The measured rate coeflicients, k, and the Langevin (collisional) rate coefficients, k,, for the reactions of ArH+ and ArD+ with H,, HD and D2. These reactions are all exothermic by -0.7eV

Reaction k (10”cm3s-‘)

300 K 80K SC’)

ArH+ + H, -+ H: + Ar 8.9 9.5 ArD+ + Hz -+ HzD+ + Ar 8.8 9.9 15.0

ArH+ + HD -+ H2D+ + Ar 8.6 9.8 ArD+ + HD + HD: + Ar 8.1 9.3 12.5

ArH+ +D2 + HD: + Ar 7.9 8.4 ArD+ + D, -+ D: + Ar 6.8 9.1 11.0

the deuterated species form H2D +, HD; , D3+. We have determined the rate coefficients for the proton (deuteron) transfer reactions of ArH+ and ArD+ with Hz, HD and D,, e.g.

ArH’ + H2 + H3f + Ar (2)

by directly injecting ArH+ and ArD+ from the ion source. The rate coefficients determined at 300 K and 80 K are listed in Table 2. Again, the k values for the various reactions display the trends expected from their reduced masses and no kinetic isotope effect is evident. This is to be expected for these reactions, which are quite exothermic (- 0.7 eV in each case). The k value at 80 K is slightly greater than the k value at 300 K for all three reactions. As has been noted for reaction 1, the rate coefficients are again always less than k,, with k, = 0.60-k, at 300K and k, N 0.75-k, at 80K for all three reactions. Proton transfer normally proceeds at the collisional rate when significantly exothermic but most other workers have also found that the k value is somewhat less than k, for these reactions [4,6,23,24].

Reactions of Ari ’ with Hz, HD and D,

The reaction of Ar,+ ’ with H, (and HD, D2) proceeds by parallel atom abstraction (reaction 3a) and switching (reaction 3b) channels, i.e.

Ar:’ + H, + ArH+ + Ar + H’ (3a)

+ ArH:’ + Ar (3b)

Values of the rate coefficient for reaction 3 determined by various workers

185

TABLE 3

The measured total rate coeflicients, k, the Langevin (collisional) rate coe&ients, k,, and the ion-product ratios for the reactions of A$ with H,, HD and D, at 300 K and 80 K. All these reactions are nearly thermoneutral (see text)

Reaction Product ratio k

(“/) (10’“cm3s-‘)

300 K 80K 300 K 80K

kc (10’0cm3s-‘)

Ar: + H2 -+ArH++Ar+H 70 52 -+ ArH: + Ar 30 48 5.0 8.3 15.0

Ar: + HD +ArH++Ar+D 40 26 -+ArD++Ar+H 40 26 -+ ArHD+ + Ar 20 48 4.3 7.1 12.5

Ar: + D2 -+ArD++Ar+D 74 50 -+ ArD: + Ar 26 50 3.6 5.2 11.0

have recently been collated by Shul et al. [7] who, in their own study, measured an overall value k, = 4.7 x 10-‘0cm3s-’ at 300K, and obtained the product ions ArH+ and ArH:. in the ratio 0.77:0.23, in good agreement with the results of this study presented in Table 3. The overall rate coefficient k, is less than collisional; this may be due to the near-thermoneutrality of the reactions and could indicate that the reactions may be slightly endothermic (see below). The rate coefficients for the two channels, k, and k,, calculated from the branching ratios and the total rate coefficient, k, , show different temperature dependences. The production of ArH+ is temperature independent, with k,(SOK) N k,(300K); but the production of ArHz’ is strongly enhanced at the lower temperature, with kb(80 K) N 4.0 k,(300 K) showing an approximate T-’ temperature dependence (assuming a linear relationship between these temperature limits). This inverse temperature dependence of k, establishes that reaction 3b is actually exothermic (although probably only slightly so), which permits a lower limit for the binding energy of Ar l H2+ * to be determined. Thus II”(Ar*H,+‘), the bond dissociation energy, must be > 1.04eV (which is consistent with the lower limit D’(Ar*H,) > 0.97 eV of Shul et al. [7]), taking D’(Ar*Ar) = 0.01 eV and the recombination energy of At-:‘, RE(Ar:‘) = 14.4eV.

No kinetic isotope effects are observed, despite the near thermoneutrality of both reaction channels (3a) and (3b), where small zero point energy differences between isotopes might be expected to be significant, since again the k scale is ,u -‘I2 . The branching ratios are the same for each reaction (within the errors) and are N 0.75:0.25 at 300 K and cx 0SO:OSO at 80 K, for the atom abstraction and switching channels. Further, there is no bias towards either product, ArH+ or ArD+, at either temperature in the HD reaction.

186

TABLE 4

The experimentally derived rate coefficients, k, and the Langevin (collisional) rate coefficients, kc, for the three reactions indicated. All three reactions are exothermic by < 1.02 eV (see text)

Reaction cm3 s-l) t:O” cm’

300 K 80 K SC’)

ArH: + Hz + ArH: + H ArHD+ + HD + ArH,D+ + D

-+ ArHD: + H ArD: + D, + ArDc + D

11.9 9.4 15.2

6.5 6.3 12.5

7.7 7.4 10.9

Reactions of ArH,f ’ (ArHD”, ArD:‘) with H2 (HD, D,): production of ArHz -like ions

ArH: is formed in the flow tube as a secondary ion by the reaction:

ArH: * + H, + ArH: + H’ (4)

The primary ion At-,+. at first reacts to produce increasing count-rates of ArH+ and ArH,f ’ as the flow of neutral reactant H, is increased. Further increase in the flow causes both these primary product ions to react with H, to produce the secondary ions H: and ArHc. The isotopic analogues ArH,D+ and ArHD: and ArD,+ are similarly formed by secondary reactions of ArHD+’ with HD, and ArDt with D2.

Production of ArH: via the termolecular reaction Hz + 2Ar + Ar- H: + Ar was first reported in a flowing afterglow experiment [25] in which argon was used as the carrier gas. It is possible, however, that the production of ArH; in the argon afterglow also proceeded first by the production of Arl ’

by the reaction Arf ’ + 2Ar, and then via reactions 3b and 4 as in the present study. ArH: has been formed also in a negative glow discharge, and its structure has been determined (from its microwave rotational spectrum) to be planar with the Ar atom bonded to the apex of the H: ion [8,26]. The unusually large dipole moment of ArH: , pD N 9 D [S], draws attention to it as an ion of particular interest.

The production of ArH: is more efficient at 80K than 300K, since the precursor ion ArH: becomes more abundant at the lower temperature (kjb is 1.5 x lo-“cm3 s-’ at 300 K, and 4.0 x 10-“cm3 s-’ at 80 K; following from Table 3). It should be stressed that ArHz is clearly stable against thermal decomposition in the SIFT at room temperature. The rate coefficients for reaction 4 and its isotopic analogues are listed in Table 4. They have been computed from the rates of formation and decay of the precursor ions ArHz .,

187

(ArHD+’ , ArD;‘) as a function of the flow of the neutral reactants, as described earlier. It can be seen that the coefticients are again less than collisional, k4 1~ 0.7 l kc at both 300 K and 80 K in all three reactions, even though the reactions are quite exothermic (by - 1 eV). The exothermicity of reaction 4 is calculated to be < 1.02 eV, by using D’(ArHl ’ ) > 1.04 eV (see previous section) and our estimate of the bond energy of ArH,f ,0.27-0.36 eV (see below). The rate coefficients for the reactions of ArHD+’ + HD and ArDz’ + D, are apparently independent of temperature (Table 4), but the ArHz + H, reaction apparently shows a temperature dependence with a higher k value at 300 K than at 80 K, although this may not be significant in view of the larger errors associated with the curve-fitting method of determin- ing the k value. No kinetic isotope effect is apparent between the production of ArH; and ArD:, since their values scale as ,L-‘/~, k z 0.75-k, in both cases.

However the k value for the HD reaction appears to be anomalously small, k z 0.5 *kc. We have no convincing explanation for this. Inspection of the raw SIFT data for this reaction shows that ArH,D+ is produced more efficiently than ArHD: at both 300K and 80K. Entropy change in the two reaction channels favours ArH,D+ production; whilst the change in rotational en- tropy in the two reaction channels is similar, the change (reduction) in the translational entropy for the channel producing ArH,D+ is - 0.29 entropy units (eu) whereas that for ArHDl production is - 0.79 eu.

Isotope exchange leads to the terminating ion ArDz via the reaction sequence

ArH,D+ 5 ArHD,f 5 ArD+ 3

It is well known that isotope exchange similarly occurs via the reaction sequence

-HD HD HD H: + H2D+ - HD: - D:

[27], indicating that this D/H exchange in the HT -like ions is not significantly inhibited by the presence of the weakly bound Ar atom.

The rate coefficient for the reaction

ArH: + H, --) H: + Ar (5)

was also determined, at 80K, in this study. The small rate coefficient ob- tained k = 5 x 10-‘2cm3s-1 suggests that the reaction may be endother- mic. Assiming that the reaction would proceed at the collisional rate (k = k,) for zero endothermicity, we can estimate the endothermicity, AE, for the reaction from the measured rate coefficient k,, by using the Arrhenius law, k, = k,exp. (- AE/kT). Thus AE = - 0.044eV for reaction 5. The bond dissociation energy Do of Ar*H: can then be obtained if that of H, *Hl is

188

known. The values for D’(H, l H:) in the literature cover quite a wide range, but there is good agreement between the recent experimental determinations by Elford [28], Beuhler et al. [29] and Hiraoka [30], who give D’(H, l H3)+ as 0.25eV, 0.29 + 0.01 eV and 0.30 f 0.01 eV, respectively, and the theoretical value of Yamagushi et al. [31] of 0.25-0.32eV. Adopting this range for D’(H: + H: + H,) and using AE = - 0.04eV from above, it follows that D’(Ar*H: ) lies within the range 0.29-0.36 eV. This is just consistent with the recent experimental value of 0.29 + 0.01 eV reported by Hiraoka and Mori [32]. The only other published value that we have found, o”(Ar*H:) = 0.36eV [33], is also just consistent with our range of D’(Ar*- H: )(although paradoxically this figure depends on a probably incorrect value for D’(H, *H: ) = 0.17 eV, from which D’(Ar - Hz ) was scaled according to the polarizabilities of H, and Ar).

CONCLUDING REMARKS

The reactions of Ar+ ’ and Ar,f * ions have been studied with H,, HD and D, at 300 K and 80 K. No significant isotope effects are observed in either the rate coefficients or the product ratios for the reactions at both temperatures. Referring only to the H, reactions as examples, the Ar+’ + H, reaction produces both ArHf and Hl’ which both react on with H,, producing H: . The Ar: ’ + H, reaction produces ArH+ and ArHz * , the latter species react- ing with H, to provide the interesting ion ArHl . The analogous reactions with HD and with D, produce the analogous deuterated ions. It is curious that the rate coefficients, k, for all the reactions studied are about 0.5-0.7 times their respective collisional rate coefficients, k,, including the k values for the exothermic proton and deuteron transfer reactions of ArH+ and ArD+ with H,, HD and D, . For exothermic proton transfer reactions involving larger (polyatomic) species the k values are invariably much closer to k, . Presumably it is the limited number of states (reduced phase space) available in the product ions into which the reaction heat can be deposited for reactions involving simple species that inhibits somewhat the reactions.

H/D exchange is observed to be facile in the reactions of ArHl -like ions with HD, the reactions ultimately proceeding to the terminating ion ArDT . H/D exchange is similarly facile in the reactions of HT-like ions, indicating that the weak bonding of the Ar atom to the H: transfer molecule does not significantly inhibit H/D exchange in the H: . The ArHT ions possess a very large permanent dipole moment, which poses the very interesting question of whether the large permanent dipole moment on the reactant ion can sig- nificantly influence the collisional (capture) rate coefficients for the reactions with molecules. It is well established that a permanent dipole moment on reactant molecules influences the capture rate coefficients with ions [34,35].

189

We are currently investigating this phenomenon; preliminary data suggest that the measured k values for the reactions of ArH,f ions with non-polar molecules do appear to exceed the kc as calculated using the Langevin theory.

ACKNOWLEDGEMENTS

We are grateful to the Science and Engineering Research Council for financial support of this work and to Kevin Giles for his assistance in taking the data on the SIFT.

REFERENCES

1 K. Giles, N.G. Adams and D. Smith, J. Phys. B, 22 (1989) 873. 2 N.G. Adams, D. Smith and E. Alge, J. Phys. B, 13 (1980) 3235. 3 D. Smith and N.G. Adams, Chem. Phys. Lett., 161 (1989) 30. 4 N.G. Adams, D.K. Bohme and E.E. Ferguson, J. Chem. Phys., 52 (1970) 5101. 5 C.B. Collins and F.W. Lee, J. Chem. Phys., 71 (1979) 184. 6 A.B. Rakshit and P. Warneck, J. Chem. Phys., 74 (1981) 2853. 7 R.J. Shul, R. Passarella, B.L. Upschulte, R.G. Keesee and A.W. Castleman, J. Chem. Phys.,

86 (1987) 4446. 8 M. Bogey, H. Bolvin, C. Demuynck and J.L. Destombes, Phys. Rev. Lett., 58 (1987) 988. 9 D. Smith and N.G. Adams, Adv. At. Mol. Phys., 24 (1981) 1.

10 K. Birkinshaw, A. Shukla, S. Howard, J. Biggerstaff and J.H. Futtrell, Int. J. Mass Spectrom. Ion Processes, 84 (1988) 283.

11 W.F. Liu and D.C. Conway, J. Chem. Phys., 62 (1975) 3070. 12 C.L. Liao, C.X. Liao and C.Y. Ng, J. Chem. Phys., 82 (1985) 5489. 13 M. Hamdan, K. Birkinshaw and N.D. Twiddy, Int. J. Mass Spectrom. Ion Processes, 62

(1984) 297. 14 A.B. Rakshit and P. Warneck, J. Chem. Phys., 73 (1980) 2673. 15 D. Smith and N.G. Adams, Phys. Rev. A, 23 (1981) 2327. 16 K.R. Ryan and L.G. Graham, J. Chem. Phys., 59 (1973) 4260. 17 R.D. Smith, D.L. Smith and J.H. Futtrell, Int. J. Mass Spectrom. Ion Phys., 19 (1976) 395. 18 K. Tanaka, J. Durup, T. Kato and I. Koyano, J. Chem. Phys., 74 (1981) 5561. 19 C. Rebrion, B.R. Rowe and J.B. Marquette, J. Chem. Phys., in press. 20 G. Gioumousis and D.P. Stevenson, J. Chem. Phys., 29 (1958) 294. 21 I. Dotan and W. Lindinger, J. Chem. Phys., 76 (1982) 4972. 22 N.G. Adams, D. Smith and E. Alge, J. Phys. B, 13 (1980) 3235. 23 A.E. Roche, M.M. Sutton, D.K. Bohme and H.I. Schiff, J. Chem. Phys., 55 (1971) 548. 24 H. Villinger, J.H. Futtrell, F. Howorka, N. Duric and W. Lindinger, J. Chem. Phys., 76

(1982) 3529. 25 F.C. Fehsenfeld, A.L. Schmeltekopf and E.E. Ferguson, J. Chem. Phys., 46 (1967) 28. 26 M. Bogey, H. Bolvin, C. Demuynck, J.L. Destombes and B.P. van Eijck, J. Chem. Phys.,

88 (1988) 4120. 27 N.G. Adams and D. Smith, in A. Fontijn and M.A.A. Clyne (Eds.), Reactions of Small

Transient Species, Academic Press, London, 1983, p. 311. 28 M.T. Elford, J. Chem. Phys., 79 (1983) 5951. 29 R.J. Beuhler, S. Ehrenson and L. Friedman, J. Chem. Phys., 79 (1983) 5982. 30 K. Hiraoka, J. Chem. Phys., 87 (1987) 4048.

190

31 Y. Yamagushi, J.F. Gaw and H.F. Schaeffer, J. Chem. Phys., 78 (1983) 4074. 32 K. Hiraoka and T. Mot-i, J. Chem. Phys., 91 (1989) 4821. 33 R.D. Poshusta and V.P. Agrawal, J. Chem. Phys., 59 (1973) 2477. 34 DC. Clary, D. Smith and N.G. Adams, Chem. Phys. Lett., 119 (1985) 320. 35 B.R. Rowe, in T.J. Millar and D.A. Williams (Eds.), Rate Coefficients in Astrochemistry,

Kluwer, Dordrecht, 1988, p. 135.