Gas phase protonation of trifluoromethyl sulfur pentafluoride

Abstract

The gas phase protonation of SF₅CF₃, a potent new greenhouse gas recently discovered in stratospheric air samples, was studied by the joint application of mass spectrometric and ab initio theoretical methods. The reaction is essentially dissociative leading to the formation of HF, CF₄ and SF₃⁺ as main fragmentation products. Consistent with collisionally activated dissociation (CAD) mass spectrometric results, theoretical calculations identified the loosely bounded ion–molecule complex [HF–SF₄–CF₃], I, as the most stable isomer on the [SF₅CF₃]H⁺ potential energy surface. The proton affinity of SF₅CF₃ estimated from FT-ICR ‘bracketing’ experiments was found to be 152.5 ± 3 kcal mol⁻¹ which agrees with the values obtained from theoretical calculations at B3LYP and CCSD(T) levels of theory, 154.0 ± 3 and 153.4 ± 3 kcal mol⁻¹, respectively. These results suggest that the basicity of SF₅CF₃ is higher than that of atmospheric cations such as H₂O⁺; they need to be considered when evaluating the lifetime of SF₅CF₃ since it can be destroyed by proton transfer reactions.

---

1. Introduction

Trifluoromethyl sulfur pentafluoride (SF₅CF₃) was recently discovered in stratospheric air samples by Sturges et al. at concentrations of about 0.12 ppt.¹ The source of SF₅CF₃ is entirely anthropogenic. Due to the similarities between the growth rate of SF₆ and SF₅CF₃ (6% per year) Sturges assumed that SF₅CF₃ could be formed as a SF₆ breakdown product in electrical equipment. However, the only known source of SF₅CF₃ is its release as a by-product during the manufacture of fluorochemicals.² The strong absorption in the infrared region³ and the long atmospheric lifetime, ranging from several hundred to a few thousand years, classify this species as one of the most powerful greenhouse gases with a global warming potential (GWP) 18 000 times that of CO₂. The estimation of its lifetime and hence of its GWP has stimulated a large number of studies on the atmospheric processes leading to its decomposition such as UV photodissociation and reaction with electrons, radicals or small ions.

It has been demonstrated that the atmospheric lifetime of SF₅CF₃ is dominated by dissociative electron attachment in the region of the ionosphere^4–6 whereas there is no SF₅CF₃ photolysis in the stratosphere. However, photodissociation is possible within the mesosphere due to the penetration of solar Lyman-α radiations.^7–10 The reactivity of SF₅CF₃ with several anions of atmospheric and industrial interest has been reported by Arnold et al.¹¹ Trifluoromethyl sulfur pentafluoride does not react with CO₃⁻ and NO₃⁻, the dominant negative ions in the lower atmosphere. A slow dissociative electron transfer reaction was observed with O₂⁻, an anion present in the atmosphere at higher altitudes. The rate constant and product distribution of the reactions between SF₅CF₃ and a large number of small cations have been measured by Kennedy and co-workers in a selected ion flow tube apparatus.^12,13 A variety of processes have been observed, most of which consist of dissociative charge transfers. In these study the only data reported on the reaction of gaseous Brønsted acid with SF₅CF₃ are those regarding the H₃O⁺ ions that were found to be unreactive. The lack of information on the SF₅CF₃ protonation is surprising, considering that proton transfer to strictly related molecules, such as SF₆, is largely dissociative. Considering the intrinsic interest of the ionic chemistry of SF₅CF₃, and taking into account the importance of determining its atmospheric lifetime, we have undertaken a comprehensive mass spectrometric and theoretical study of the gas phase protonation of SF₅CF₃, aimed, in particular, at evaluating its unknown proton affinity (PA). To validate the experimental method used the PA of SF₆ was also re-evaluated.^14,15

Finally, in order to investigate the possible formation of SF₅CF₃ in electrical devices through ionic or radical reactions involving SF₆ and fluoropolymers, a number of gaseous mixtures containing sulfur hexafluoride and fluorocarbons such as CF₄, C₂F₆ and C₃F₈ were subjected to corona discharge processes.

2. Experimental

All the gases used were purchased from Matheson Gas Products Inc. with a stated purity of 99.9 mol% except for SF₅CF₃ that was obtained from Fluorochem Ltd with a stated purity of 96.0 mol%.

2.1. Corona discharge reactor

The source of the corona discharge was an Edwards ST200k spark tester capable of delivering a 0–50 kV pulse of 200 kHz of radio-frequency. The discharge cell electrodes consist of a stainless steel cylinder and a copper wire (3.5 × 10⁻⁴ m diameter) mounted along the cylinder’s central axis. The cylinder is 0.10 m long with a 3.0 × 10⁻³ m wall thickness and has an inside diameter of 2.5 × 10⁻² m. A glass window on one side permits to view the internal of the reactor to ensure that it is operating with a pure corona plasma. Two Teflon fittings are used as anchors of the central electrode on each end of the reactor. The inlet and the outlet are placed in the opposite side respect to the glass window and are connected directly to the neutral reactant gas cylinder and the mass spectrometer, respectively. Typical operating voltage of the reactor was 8–10 kV.

2.2. Mass spectrometric experiments

Triple quadrupole mass spectrometric experiments were performed with a TSQ 700 instrument from Thermofinnigan Ltd. The ions generated in the chemical ionization (CI) source were driven into the collision cell, actually a RF-only hexapole, containing the neutral reagent. The collisionally activated dissociation (CAD) spectra were recorded utilizing Ar as the target gas at pressures up to 1 × 10⁻⁵ Torr and at collision energies ranging from 0 to 25 eV (laboratory frame). The charged products were analyzed with the third quadrupole, scanned at a frequency of 150 amu s⁻¹, accumulating about 150 scans for each run.

FT-ICR measurements were performed using an Apex TM 47e spectrometer from Bruker Spectrospin AG equipped with an external ion source operating in the CI mode. The ions, generated in the external source, were transferred into the resonance cell, isolated by broad band and ‘single shot’ ejection pulses and thermalized by collisions with Ar which was introduced in the cell by a pulsed valve. The pressure of the neutral reactant introduced in the cell, ranging from 1 × 10⁻⁸ to 1 × 10⁻⁷ Torr, was measured by a Bayard–Alpert ionization gauge whose reading was corrected for the relative sensitivity to the various gases used according to standard procedures.¹⁶ The pseudo-first-order rate constants were obtained by plotting the logarithm of the BH⁺_t/BH⁺_t=0 ionic intensity ratio as a function of the reaction time. Then the bimolecular rate constants were determined from the number density of the neutral molecules deducted from the gas pressure. Average dipole orientation (ADO) collision rate constants, K_ADO, were calculated as described by Su and Bowers.¹⁷ Reaction efficiencies are the ratios of experimental rate constant, K_exp, to the collision rate constant, K_ADO. The uncertainty of each rate constant is estimated to be of about 30%.

2.3. Computational details

Density functional theory, using the hybrid¹⁸ B3LYP functional¹⁹ has been used to localize the stationary points of the investigated systems and to evaluate the vibrational frequencies. Single point energy calculations at the optimized geometries were performed using the coupled-cluster single and double excitation method²⁰ with a perturbational estimate of the triple excitations approach CCSD(T).²¹ Zero point energy corrections evaluated at B3LYP level were added to the CCSD(T) energies. The 0 total energies of the species of interest were corrected to 298 K by adding translational, rotational and vibrational contributions. The absolute entropies were calculated by using standard statistical-mechanistic procedures from scaled harmonic frequencies and moments of inertia relative to B3LYP/6-311++G(d,p) optimized geometries. The 6-311++G(d,p) basis set²² has been used. All calculations were performed using Gaussian 03²³ and MOLPRO.²⁴

3. Results

3.1. Protonation of SF₅CF₃

Protonation of SF₅CF₃ occurs in the sources of TQ and FT-ICR mass spectrometers under chemical ionization conditions according to the general reaction (1) where B = H₂, CH₄ or CO.

SF₅CF₃ + BH⁺ → SF₅CF₃H⁺ + B (1)

A typical CH₄/CI spectrum of SF₅CF₃ is shown in Fig. 1. Extensive decomposition of the [SF₅CF₃]H⁺ ions readily occurs in the ion source as suggested by the extremely low intensity of the MH⁺ ion at m/z 197 and the higher intensity of the fragment ions CF₃⁺ and SF₃⁺ at m/z 69 and 89, respectively. A similar spectrum was obtained using the COH⁺ ions produced by the chemical ionization of 1 ∶ 1 H₂/CO mixture as protonating species. Only slight traces of protonated SF₅CF₃ were obtained when H₃⁺ ions were used. Besides the MH⁺ species at m/z 197, the fragment ion at m/z 177, corresponding to the loss of an HF molecule, is characteristic of the chemical ionization of SF₅CF₃ since it has never been observed either in electron impact or photoionization conditions. The formation of the SF₅⁺ ion at m/z 127 may be partially due to the presence of SF₆ as an impurity in the SF₅CF₃ sample.

Fig. 1 CH₄/CI spectrum of trifluoromethyl sulfur pentafluoride.

Proton transfer reaction was also studied in the collision cell of the TQ mass spectrometer by observing the reactivity of mass selected protonating species such as CH₅⁺, COH⁺ and H₂O⁺ towards SF₅CF₃ at nominal collision energy ranging from 0 to 8 eV (laboratory frame). Under the low pressure conditions of the collision cell no [SF₅CF₃]H⁺ ions at m/z 197 and fragment ions at m/z 177 were observed. Furthermore, the SF₃⁺ ion at m/z 89 dominated the spectra at collision energy zero and decreased with increasing collision energy.

Experiments performed by mass selecting the SF₅CF₃H⁺ ions allow reaction (1) to be studied in the opposite direction. The [SF₅CF₃]H⁺ ions do not behave as a protonating species since proton transfer has never been detected when bases such as COS, CH₃Cl and H₂O are introduced in the cell. The only process observed was the decomposition into the fragment ions at m/z 69, 89 and 177. The trend of the intensities of these species increasing the collision energy resembles that observed in the CAD spectra (vide infra).

3.2. Structural characterization of [SF₅CF₃]H⁺ ions

Two protomers of SF₅CF₃, I and II, are conceivably formed through reaction (1).

Collisionally activated dissociation (CAD) mass spectrometry has been used to obtain structural information on the [SF₅CF₃]H⁺ ionic population obtained from CH₄/CI of SF₅CF₃. The low-energy CAD spectrum of the ions at m/z 197, recorded at nominal collision energies ranging from 0 to 25 eV (laboratory frame), is reported in Fig. 2. At low collision energies the spectrum is dominated by the fragment at m/z 177, corresponding to the loss of the HF molecule, which unambiguously supports the occurrence of protonation on a fluorine atom. At higher collision energies the intensity of the ion at m/z 177 decreases while the CF₃⁺ ion intensity at m/z 69 simultaneously increases, becoming the most important dissociation channel at 25 eV. Interestingly, the SF₃⁺ fragment at m/z 89 has its maximum intensity at nominal collision energy zero and decreases as the collision energy increases.

Fig. 2 Energy resolved CAD spectrum of [SF₅CF₃]H⁺ ions obtained from reaction (1).

3.3. The PA of SF₅CF₃

Protonated SF₅CF₃ generated in the external CI source of the FT-ICR mass spectrometer was isolated by soft ejection techniques in the resonance cell. At the low pressure conditions of the ICR cell this species does not react with added bases but undergoes exclusively a fast decomposition into CF₃⁺ and SF₃⁺. Owing to the prone dissociation of the [SF₅CF₃]H⁺ ions and their inability to behave as protonating species, the only viable procedure for measuring the basicity of SF₅CF₃ is the ‘bracketing method’.^25,26 According to this method, the PA of gaseous species can be estimated by measuring the efficiency of proton transfer reaction promoted by reference bases (BH⁺) of known PA. For exothermic proton transfer reactions the efficiency is high, for strongly endothermic processes it drops to zero, whereas for slightly endothermic reactions it is low, but still measurable.

In order to estimate the PA of SF₅CF₃, BH⁺ ions, generated in the external ion source by CH₄ or H₂ chemical ionization, were isolated in the resonance cell, thermalized by collision with Ar and then allowed to react with SF₅CF₃. The efficiencies of the reaction (1) obtained from the linear decrement of the BH⁺ intensity over time are listed in Table 1. All the PA values used are in accordance with the most recent NIST database²⁷ in order to make use of a self-consistent scale based on a single anchored value.

Table 1 ICR reactivity of BH⁺ ions towards SF₅CF₃

					Product ions (r.i.%)^c
BH^+	PA(B)/kcal mol^−1	k ADO ^a	k exp ^a	RE (%)^b	SF3^+	CF3^+
a Units: 10^−9 cm^3 s^−1 mol^−1. b RE, reaction efficiency. c Mean relative intensities taken at reaction time between 0.1–1 s.
CO2H^+	129.2	1.12	1.21	100	32.0	68.0
N2OH^+	137.5	1.12	1.26	100	38.0	62.0
COH^+	142.0	1.31	1.30	100	40.4	59.6
COSH^+	150.2	0.99	0.36	36.4	65.0	35.0
CH3ClH^+	154.7	1.06	0.33	31.1	63.2	36.8
C2H5^+	162.6	1.33	0.067	5.0	100

---

A typical time profile of the ionic intensities from the reaction of N₂OH⁺ and COSH⁺ ions with SF₅CF₃ is shown in Fig. 3.

Fig. 3 Time profiles of the ionic intensities in the reaction of [N₂O]H⁺ (a) and [COS]H⁺ (b) ions with SF₅CF₃ under FT-ICR conditions.

Under the low pressure conditions of the FT-ICR cell no [SF₅CF₃]H⁺ ions at m/z 197 or fragment ions at m/z 177 were detected. The occurrence of the proton transfer reaction was probed by the formation of the SF₃⁺and CF₃⁺ fragment ions at m/z 69 and 89, respectively. No fragmentations were observed when ions such as H₃O⁺ were used. Proton transfer from C₂H₅⁺ ions (PA C₂H₄ = 162.6 kcal mol⁻¹) leads to the formation of traces of the SF₃⁺ ion with a very low efficiency. An efficiency approaching one was measured when protonation was accomplished using acids such as COH⁺ (PA CO = 142.0 kcal mol⁻¹) and N₂OH⁺ (PA N₂O = 137.5 kcal mol⁻¹). Besides SF₃⁺, the proton transfer from these acids leads also to the formation of the CF₃⁺ fragment ion. A proton transfer efficiency around 50% was observed with COSH⁺ (PA COS = 150.2 kcal mol⁻¹) and with CH₃ClH⁺ (PA CH₃Cl = 154.7 kcal mol⁻¹). In these cases SF₃⁺ was the major reaction product observed whereas minor amount of CF₃⁺ were formed. It is worth to note from the time profiles of the ionic intensities shown in Fig. 3 the decrement of the CF₃⁺ ion intensity at long reaction time. This behaviour, common to all the investigated proton transfer processes, can be explained by the reaction of CF₃⁺ with SF₅CF₃ giving SF₃⁺ previously described by Kennedy et al.¹³

On the basis of the proton transfer efficiencies, it is reasonable to assume that the proton affinity of SF₅CF₃ is intermediate between that of CH₃Cl and COS, which defines a PA interval from 154.7 to 150.2 kcal mol⁻¹. In evaluating the uncertainty range of this measurement, several sources of errors, and their propagation, need to be considered. Indeed, apart from the error associated with the tabulated PA values of the reference bases used, a significant contribution could come from the measurement of the proton transfer efficiencies deriving from the dissociation of the [SF₅CF₃]H⁺ ions. By taking into account the various sources of error, a reasonable estimate of the overall uncertainty of the experimental PA of SF₅CF₃ is of the order of ±3 kcal mol⁻¹. Given the limited accuracy of these measurements, the PA of SF₅CF₃ can be estimated as 152.5 ± 3 kcal mol⁻¹.

To verify the method used, we re-evaluated the PA of SF₆ by applying the same procedure. As is well known, also in this case proton transfer leads to the fast decomposition of protonated sulfur hexafluoride into the SF₅⁺ fragment ion. The efficiencies of the reactions of selected BH⁺ ions toward SF₆ are reported in Table 2. The data are in good agreement with the value obtained by Bohme et al.¹⁴ in their FA/SIFT studies. The PA of SF₆ can be considered as that of N₂O (PA = 137.5 kcal mol⁻¹), in accordance with the existing literature data probing the validity of the experimental method used to derive the PA of SF₅CF₃.

Table 2 ICR reactivity of BH⁺ ions towards SF₆

BH^+	PA(B)/kcal mol^−1	K ADO ^a	K exp ^a	RE (%)
a Units: 10^−9 cm^3 s^−1 mol^−1.
CO2H^+	129.2	0.85	1.11	100
CH5^+	129.9	1.27	1.31	100
N2OH^+	137.5	0.85	0.30	35.3
COH^+	142.0	1.0	0.17	17.0

---

3.4. Theoretical calculations

In order to rationalize the experimental results, theoretical calculations have been performed with an approach based on the density functional theory using the hybrid B3LYP functional. Additional calculations have also been carried out using the coupled-cluster single and double excitation method with a perturbational estimate of the triple excitations CCSD(T) approach.

Two minima, corresponding to the structures I and II were found on the potential energy surface of the [SF₅CF₃]H⁺ ions. Their geometrical parameters are illustrated in Fig. 4 together with those of neutral SF₅CF₃. The potential energy diagram of the [SF₅CF₃]H⁺ ions is reported in Fig. 5. Protonation at one of the fluorine atoms of the SF₅ group gives rise to the most stable isomer I. Considering the very long S–FH bond distance, 3.329 Å, protomer I can be viewed as a loosely bound ion–molecule complex between SF₄CF₃⁺ and HF with a dissociation energy computed to be 17.3 kcal mol⁻¹ at B3LYP level. Also the S–C bond in the SF₄CF₃⁺ fragment was found to be relatively elongated with respect to neutral SF₅CF₃.

Fig. 4 B3LYP/6-311++G(d,p) optimized geometries of the [SF₅CF₃]H⁺ ions. Bond lengths in ångströms and bond angles in degrees.

Fig. 5 Theoretical potential energy diagram of the [SF₅CF₃]H⁺ ions. All thermodynamic parameter are in kcal mol⁻¹.

The dissociation of this ion into its constituents, CF₃⁺ and SF₄ is exothermic by 19.9 kcal mol⁻¹. However the saddle for this dissociation is so low, that it disappears once we include the zero point vibrational energy correction. It is worth noting that the easier fragmentation channel of isomer I is the dissociation into SF₃⁺ HF and CF₄, a process computed to be largely exothermic (49.3 kcal mol⁻¹). Unfortunately, we were not able to localize the transition states for this process, despite our many attempts. The potential energy surface of the [HF–SF₄–CF₃]⁺ ions is very flat as the system is composed of three fragments electrostatically interacting among them. However, we can hypothesize that the transition state for the fragmentation of I into SF₃⁺, CF₄ and HF is close in energy terms to the starting minimum.

The proton affinity of SF₅CF₃ to the fluorine atoms of the SF₅ group is computed to be 154.0 ± 3 and 153.4 ± 3 kcal mol⁻¹, at the B3LYP and CCSD(T) levels of theory, respectively.

Protonation at the fluorine atoms of the CF₃ group gives rise to the less stable isomer II situated 33.7 and 26.7 kcal mol⁻¹ higher in energy than ion I at the B3LYP and CCSD(T) levels of theory, respectively. The interconversion of II into I is characterized by an activation energy of 8.9 kcal mol⁻¹ at the B3LYP level and 13.6 kcal mol⁻¹ at CCSD(T) level. Considering the relatively long C–FH bond distance, 2.286 Å, also this species can be viewed as a loosely bound ion–molecule complex between SF₅CF₂⁺ and HF characterized by a dissociation energy of 12.0 kcal mol⁻¹ at B3LYP level. Protonation on the CF₃ group results in tighter S–F and C–F bonds, whereas the S–C bond length is almost completely unaffected. The dissociation energy of SF₅CF₂⁺ into the SF₅⁺ and CF₂ moieties amounts to 42.3 kcal mol⁻¹.

3.5. Is SF₅CF₃ generated in electrical devices?

In order to test the formation of SF₅CF₃, from the recombination of SF₅ and CF₃ radicals generated as breakdown products in high-voltage equipment containing SF₆ and fluoropolymers, several mixtures of SF₆ and fluorocarbons such as CF₄, C₂F₆ and C₃F₈ were subjected to corona discharge processes. The sparked gaseous mixtures were introduced into the source of the mass spectrometer operating in CH₄/CI mode. In these conditions, the possible formation of neutral SF₅CF₃ can be proved by the detection of its peculiar ions at m/z 197 and 177. No traces of these species were observed in any of the mixtures investigated.

4. Discussion

Protonation of SF₅CF₃ was first accomplished in the CI source of both TQ and FT-ICR mass spectrometers using gaseous Brønsted acids such as CH₅⁺, COH⁺ and H₃⁺. In these conditions appreciable amounts of the [SF₅CF₃]H⁺ ions at m/z 197 were detected, although the proton transfer reaction was largely dissociative, as it is evident from the extensive formation of the fragment ions CF₃⁺ and SF₃⁺. Furthermore, a low intensity of the fragment ion at m/z 177, corresponding to the loss of an HF molecule, was observed, in contrast to the protonation of related molecules such as SF₆ and CF₄, leading almost entirely to the formation of the SF₅⁺ and CF₃⁺ fragment ions. The SF₃⁺ and CF₃⁺ fragments represent the main ionic products of the reaction of mass selected BH⁺ ions toward SF₅CF₃ in the low pressure conditions of both the TQ and FT-ICR cells and can be considered as indicative of the occurrence of the proton transfer reaction. The mutually supporting CAD and theoretical results suggest the formation from reaction 1 of an ionic population characterized by the [HF–SF₄–CF₃]⁺ connectivity characteristic of protomer I. The theoretically calculated potential energy surface shown in Fig. 5 allows the experimental results to be rationalized. The fragmentation into the CF₃⁺ and SF₃⁺ ions is peculiar to protomer I and demonstrates its formation. The SF₄CF₃⁺ ion, at m/z = 177, is formed from isomer I through the endothermic loss of an HF molecule and undergoes a subsequent fragmentation into the CF₃⁺ ions in a barrierless process. In accordance with the theoretical results, the energy resolved CAD spectra of the [SF₅CF₃]H⁺ ions displayed in Fig. 2 show that the SF₄CF₃⁺ ion intensity rises increasing the collision energies and that the CF₃⁺ fragment arises from SF₄CF₃⁺ decomposition since its intensity increases as its precursor decreases. In addition, the low intensity of the SF₄CF₃⁺ ions in the CI spectra is due to their easy decomposition into CF₃⁺, a process for which no appreciable dissociation barrier was found through theoretical calculations. Even the dissociation channel leading to the SF₃⁺ fragment is a proof of isomer I formation. Considering the long S–FH and S–C bond distances, protomer I can be considered as a complex in which the HF, SF₄ and CF₃ moieties are electrostatically bounded to each other. Taking into account the weak S–C interaction, the [HF–SF₄–CF₃]⁺ complex I can easily dissociates into the HF, SF₃⁺ and CF₄ fragments probably throughout the formation of an ion–molecule complex in which the SF₄ and CF₃ groups are coordinated by a fluorine atom. Although all theoretical attempts to find an energy barrier for this process failed, it must be assumed to exist, albeit lying close to the starting minimum, in view of the changes in the complex coordination and experimental evidence indicating the formation and the isolation of stable [SF₅CF₃]H⁺ ions. Finally, three factors allow the existence of protomer II to be excluded: (i) proton transfer to the fluorine atom of the CF₃ group from almost all the gaseous Brønsted acids used is thermodynamically forbidden, (ii) isomer II if eventually formed can undergo intermolecular and/or intramolecular isomerization into I, considering their thermodynamic stability and the relatively low interconversion barrier, and (iii) no SF₅⁺ fragment ions, characteristic of isomer II decomposition were observed in the CAD spectra.

The PA of SF₅CF₃ can be considered intermediate between that of COS and CH₃Cl on the basis of the proton transfer reaction efficiency and of energetic considerations concerning the [SF₅CF₃]H⁺ ion fragmentation. In fact, proton transfer to SF₅CF₃ from these bases has an efficiency of around 50% and is almost thermoneutral since leads predominantly to fragmentation into the SF₃⁺ ion, characterized by a very low dissociation barrier. Conversely, the CF₃⁺ fragment, the formation of which requires at least 17.3 kcal mol⁻¹, appears in exothermic proton transfer reactions having a 100% efficiency as in the case of COH⁺ and N₂OH⁺. The experimental PA value of 152.5 ± 3 kcal mol⁻¹ agrees with the value computed at the B3LYP and CCSD(T) level of theory, amounting to 154.0 ± 3 and 153.4 ± 3 kcal mol⁻¹, respectively. The PA values of SF₅CF₃ and of the OH radical (141.8 kcal mol⁻¹) account for the unexpected formation of an intense SF₃⁺ ion in the reaction between H₂O⁺ and SF₅CF₃ observed in a selected ion flow tube apparatus by Kennedy et al.¹³ These authors, while studying the reactivity of several cations toward SF₅CF₃, observed the anomalous reactivity of the H₂O⁺ ion with respect to the other cations investigated, which react by charge transfer and lead to the formation of CF₃⁺ as the major product. Since charge transfer from H₂O⁺ to SF₅CF₃ is not allowed, they hypothesized a different reaction mechanism in which bonds are broken and formed. Our results confirm their assumption that H₂O⁺ ions transfer the proton to SF₅CF₃ leading to its fast decomposition into SF₃⁺ ions.

As to the origin of SF₅CF₃ in the atmosphere, the hypothesis of its formation in electrical devices through ionic or radical reactions involving SF₆ and fluoropolymers can be ruled out. No traces of the ions at m/z 197 and 177, which are peculiar to the chemical ionization of SF₅CF₃, were observed in any of the mixtures containing sulfur hexafluoride and fluorocarbons submitted to corona discharge processes.

5. Conclusions

The gas phase protonation of a potent new greenhouse gas, SF₅CF₃, was studied for the first time by the joint application of mass spectrometric and ab initio theoretical methods. The reaction is essentially dissociative and leads to the formation of the SF₃⁺ and CF₃⁺ ions as the main fragmentation products. Experimental and theoretical results provide strong evidence for the existence in the gas phase of a single fluorine-protonated isomer of SF₅CF₃. The most stable isomer I can be viewed as a loosely bounded ion–molecule complex [HF–SF₄–CF₃]⁺ and, accordingly, does not behave in FT-ICR and TQ conditions as a Brønsted acid but undergoes fast dissociation exclusively.

The mutual application of theoretical and experimental methods allowed the evaluation of the unknown PA of trifluorometyl sulfur pentafluoride, from which the best theoretical estimate of 153.4 ± 3 kcal mol⁻¹ at the CCSD(T) level is comparable to the experimental value of 152.5 ± 3 kcal mol⁻¹, obtained by utilizing the FT-ICR ‘bracketing’ method.

These results indicate that gas phase protonation of SF₅CF₃ by cations such as H₂O⁺ represents a reaction that can be considered in evaluating its atmospheric lifetime since this could be responsible for its fast decomposition.

Finally, no evidence for the genesis of neutral SF₅CF₃ in the electrical equipment was found.

Acknowledgements

This work was carried out with the financial support of the University of Rome “La Sapienza” and the Consiglio Nazionale delle Ricerche (CNR).

References

1. W. T. Sturges, T. J. Wallington, M. D. Hurley, K. P. Shine, K. Shira, A. Engel, D. E. Oram, S. A. Penkett, R. Mulvaney and C. A. M. Brenninkmeijer, Science, 2000, 289, 611.
2. M. A. Santoro, Science, 2000, 290, 935.
3. O. J. Nielsen, F. M. Nicolaisen, C. Bacher, M. D. Hurley, T. J. Wallington and K. P. Shine, Atmos. Environ., 2002, 36, 1237.
4. R. A. Kennedy and C. A. Mayhew, Int. J. Mass Spectrom., 2001, 206, i.
5. T. M. Miller, S. T. Arnold, A. A. Viggiano and W. B. Knighton, J. Chem. Phys., 2002, 116, 6021.
6. R. Balog, M. Stano, P. Limão-Vieira, C. König, I. Bald, N. J. Mason and E. Illenberger, J. Chem. Phys., 2003, 119, 10396.
7. R. Y. L. Chim, R. A. Kennedy, R. P. Tuckett, W. Zhou, G. K. Jarvis, D. J. Collins and P. A. Hatherly, J. Phys. Chem. A, 2001, 105, 8403.
8. K. Takahashi, T. Nakayama, Y. Matsumi, S. Solomon, T. Gejo, E. Shigemasa and T. J. Wallington, Geophys. Res. Lett., 2002, 29, 7–11.
9. R. Y. L. Chim, R. A. Kennedy and R. P. Tuckett, Chem. Phys. Lett., 2003, 367, 697.
10. T. Ibuki, Y. Shimada, S. Nagaoka, A. Fujii, M. Hino, T. Kakiuchi, K. Okada, K. Tabayashi, T. Matsudo, Y. Yamana, I. H. Suzuki and Y. Tamenori, Chem. Phys. Lett., 2004, 392, 303.
11. S. T. Arnold, T. M. Miller, A. A. Viggiano and C. A. Mayhew, Int. J. Mass Spectrom., 2003, 223–224, 403.
12. C. Atterbury, R. A. Kennedy, C. A. Mayhew and R. P. Tuckett, Phys. Chem. Chem. Phys., 2001, 3, 1949–1953.
13. C. Atterbury, A. D. J. Critchley, R. A. Kennedy, C. A. Mayhew and R. P. Tuckett, Phys. Chem. Chem. Phys., 2002, 4, 2206.
14. G. I. Mackay, H. I. Schiff and D. K. Bohme, Int. J. Mass Spectrom. Ion Processes, 1992, 117, 387.
15. D. R. Latimer and M. A. Smith, J. Chem. Phys., 1994, 101, 3410.
16. J. E. Bartmess and R. M. Georgiadis, Vacuum, 1983, 33, 149.
17. T. Su and M. T. Bowers, Int. J. Mass Spectrom. Ion Phys., 1973, 12, 347.
18. A. D. Becke, J. Chem. Phys., 1993, 98, 5648.
19. P. J. Stevens, F. J. Devlin, C. F. Chabalowski and M. J. Frisch, J. Chem. Phys., 1994, 98, 11623.
20. R. J. Bartlett, Annu. Rev. Phys. Chem., 1981, 32, 359.
21. K. Raghavachari, G. W. Trucks, J. A. Pople and M. Head-Gordon, Chem. Phys. Lett., 1989, 157, 479.
22. M. J. Frisch, J. A. Pople and J. S. Binkley, J. Chem. Phys., 1984, 80, 3265 , and references therein.
23. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgomery, Jr, T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, C. Gonzalez and J. A. Pople, GAUSSIAN 03, (Revision B.04), Gaussian Inc., Pittsburgh, PA, 2003.
24. R. D. Amos, A. Bernhardsson, A. Berning, P. Celani, D. L. Cooper, M. J. O. Deegan, A. J. Dobbyn, F. Eckert, C. Hampel, G. Hetzer, P. J. Knowles, T. Korona, R. Lindh, A. W. Lloyd, S. J. McNicholas, F. R. Manby, W. Meyer, M. E. Mura, A. Nicklass, P. Palmieri, R. Pitzer, G. Rauhut, M. Schütz, U. Schumann, H. Stoll, A. J. Stone, R. Tarroni, T. Thorsteinsson and H.-J. Werner, MOLPRO, a package of ab initio programs designed by H.-J. Werner and P. J. Knowles, Version 2002.1, 2002.
25. S. G. Lias, J. F. Liebman and R. D. Levin, J. Phys. Chem. Ref. Data, 1984, 13, 695.
26. S. G. Lias, J. E. Bartmess, J. F. Liebman, J. H. Holmes, R. D. Levin and W. G. Mallard, J. Phys. Chem. Ref. Data Suppl., 1998, 17(1), 1–861.
27. NIST Chemistry WebBook; NIST standard Reference Database No. 69 – February 2000 release; data collection of the National Institute of Standards and Technology (http://webbook.nist.gov).

---