Technical Comments

Comment on “Activation of methane to CH3+: A selective industrial route to methanesulfonic acid”

See allHide authors and affiliations

Science  10 May 2019:
Vol. 364, Issue 6440, eaax7083
DOI: 10.1126/science.aax7083


Díaz-Urrutia and Ott (Reports, 22 March 2019, p. 1326) report a selective conversion of methane to methanesulfonic acid that is proposed to occur by a cationic chain reaction in which CH3+ adds to sulfur trioxide (SO3) to form CH3–S(O)2O+. This mechanism is not plausible because of the solvent reactivity of CH3+, the non-nucleophilicity of the sulfur atom of SO3, and the high energy of CH3–S(O)2O+.

The direct synthesis of methanesulfonic acid (MSA) from methane and sulfur trioxide (SO3) is a potentially high-value but challenging transformation. Older approaches to this conversion involved either metal-catalyzed reactions or catalysis by peroxo salts, generally in fuming sulfuric acid (oleum) solutions, in reactions that are thought to occur by free-radical chain mechanisms (1, 2). Recently, Díaz-Urrutia and Ott reported the highly selective formation of MSA from methane in oleum catalyzed by sulfonyl peroxide derivatives, and it was proposed that the high yields and selectivity observed were the result of a novel mechanism involving CH3+ in a cationic chain reaction (3). In the proposed mechanism (Fig. 1), the protonated catalyst (1) carries out a hydride abstraction from methane to give the chain-carrying CH3+. The CH3+ is then proposed to add to an S=O bond of SO3 at the sulfur (as in 2) to produce CH3–S(O)2O+ (3). The CH3–S(O)2O+ would then abstract a hydride from methane to afford the MSA and CH3+ to continue the chain. Notably, no methyl bisulfate (MBS; CH3OSO3H) could be observed in this reaction, and the success of as little as 0.1% catalyst was indicative of very long chains.

Fig. 1 Mechanistic proposal of Díaz-Urrutia and Ott.

The proposed mechanism is supplemented here by three of the resonance structures for SO3.

We describe here why this mechanism is unreasonable in a series of ways based on literature precedent and standard chemical precepts. To add quantitative insight into the energetic infeasibility of the process, computational studies were carried out on the key structures in CCSD(T)/aug-cc-pVTZ//M06-2X/6-31+G(d,p) calculations including an SMD implicit solvent correction, supplemented by CAS(10,10)-NEVPT2 for open-shell structures of 3 (4).

A first concern is that CH3+ would be too reactive with the solvent to be an intermediate in a chain reaction. Strongly stabilized carbocations have long been directly observed in sulfuric acid solutions, but the range of observable cations is limited by the basicity and nucleophilicity of sulfuric acid. At an extreme, the marginally observable cumyl cation PhC+(CH3)2 requires 30% oleum (5). CH3+ is in contrast an intrinsically high-energy structure; gas-phase hydride affinities place CH3+ at 94 kcal/mol above the cumyl cation (68). This high energy fits with solution observations. Olah et al. were unable to observe CH3+ under the most stringently non-nucleophilic superacid conditions (9). From an extrapolated pKR value, CH3+ is less favorable to form in solution than the cumyl cation by ~30 orders of magnitude (10). If CH3+ were formed in an oleum solution, its reaction with the sulfuric acid to form CH3OS(O)(OH)2+ (4, Fig. 2) is computationally predicted by the CCSD(T)/SMD//M06-2X calculations to be barrierless in potential energy and downhill in free energy by 29.5 kcal/mol. Proton transfer from CH3OS(O)(OH)2+ to solvent would afford MBS, an unreactive thermodynamic sink in these reactions. For a successful chain reaction requiring little catalyst, the CH3+ would have to react ≫100 times faster with the less-basic SO3 than it undergoes the barrierless reaction with solvent.

Fig. 2 Calculated energies for key structures.

The free energies shown are based on CCSD(T)/aug-cc-pVTZ single-point energies and SMD explicit solvent corrections for gas-phase structures optimized in M06-2X/6-31+G** calculations, with a standard state of 1 M and 25°C. An unrestricted broken-symmetry basis was used for 3-open shell. CAS(10,10)-NEVPT2 calculations place 3-open shell at 25.1 kcal/mol above 3-Cs.

A second concern is that SO3 is not nucleophilic at its sulfur atom. Although SO3 reacts widely with nucleophiles and free radicals, the literature does not contain any examples of SO3 reacting with electrophiles at the sulfur atom. The non-nucleophilicity of the sulfur atom is readily understandable from simple resonance considerations, as significant contributing resonance structures (Fig. 1) place a positive charge on the sulfur but no valid resonance structure places either a negative charge or a lone pair of electrons on the sulfur.

An interrelated third concern is that the CH3–S(O)2O+ product of the proposed electrophilic addition of CH3+ to the sulfur atom is an oxylium ion (a monovalent oxygen cation). Oxylium ions are exceedingly high-energy structures; hydride affinities place HO+ 98 kcal/mol above the high-energy CH3+ (6). From this, it would be expected that the CH3–S(O)2O+ would be higher in energy than the separate CH3+ and SO3. Calculations support this expectation. Figure 2 shows the calculated energetics for a series of relevant structures. The closed-shell symmetrical 3-C3V structure for CH3–S(O)2O+ is not a minimum on the potential energy surface, and it is predicted to be 61.6 kcal/mol above the starting CH3+ / SO3. An asymmetric open-shell form of 3 (3-open shell) was an energy minimum but remained extremely high in energy in both CCSD(T) and CAS(10,10)-NEVPT2 calculations. The lowest-energy structure arising from CH3+ binding at the sulfur of SO3 is one in which two oxygens have bonded to form a strained dioxathiirane (3-Cs). This structure avoids placing a formal positive charge on an oxygen, but it is still a prohibitive 37.6 kcal/mol above the CH3+ / SO3. The only thermodynamically feasible reaction of CH3+ with SO3 is bond formation with an oxygen atom to afford 5. Díaz-Urrutia and Ott noted that bond formation at the oxygen atom could not account for the formation of MSA (3), and the formation of 5 is in any case predicted to be less favorable than reaction of CH3+ with the sulfuric acid to form 4, by 15.5 kcal/mol. The preferred methylation of the solvent fits with the known 27 kcal/mol greater proton affinity of H2SO4 over SO3 (7).

These considerations preclude any possibility of the MSA being formed by the proposed chain process involving CH3+ and CH3–S(O)2O+. The actual mechanism for this impressive transformation remains unknown. Díaz-Urrutia and Ott excluded a free-radical mechanism based largely on observations in the presence of free-radical inhibitors, but we note that a uniquely rapid free-radical step may potentially complicate experimental observations. In part because of a favorable quadrupolar interaction of a methyl radical (CH3•) with SO3, the two are predicted to form a free energy–favored face-to-face complex (at –2.0 kcal/mol versus the separate CH3• / SO3). The barrier for the subsequent addition via transition state 6 is then minimal. As a result, the CH3• / SO3 reaction would occur at approximately a diffusion-controlled rate and, most notably, as quickly as CH3• could react with any free-radical inhibitor. Under such circumstances, the interpretation of observations in the presence of inhibitors requires great care.


Acknowledgments: Funding: Supported by NIH grant GM-45617. Author contributions: Both authors conceived the study and carried out the computations. D.A.S. wrote the first draft of the manuscript and both authors worked on the final version. Competing interests: The authors declare no competing interests. Data and materials availability: The detailed computational methods and the computed structures and energies are available at the Harvard Dataverse (11).
View Abstract

Navigate This Article