Subscriber access provided by Caltech Library C: Surfaces, Interfaces, Porous Materials, and Catalysis DFT Mechanistic Study of Methane Mono-Esterification by Hypervalent Iodine Alkane Oxidation Process Ross Fu, Robert J. Nielsen, Nichole Schwartz Liebov, William A. Goddard, T. Brent Gunnoe, and John T. Groves J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.9b04239 • Publication Date (Web): 28 May 2019 Downloaded from http://pubs.acs.org on May 28, 2019

Just Accepted

“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties. Page 1 of 12 The Journal of Physical Chemistry

1 2 3 4 5 6 7 DFT Mechanistic Study of Methane Mono-Esterification by 8 Hypervalent Iodine Alkane Oxidation Process 9 10 Ross Fu†, Robert J. Nielsen†, Nichole S. Liebov‡, William A. Goddard III†*, T. Brent Gunnoe‡, John T. 11 Groves§ 12 13 †Materials and Process Simulation Center (139-74), California Institute of Technology, Pasadena, CA 91125, USA 14 15 ‡Department of Chemistry, University of Virginia, Charlottesville, VA 22904, USA 16 17 §Department of Chemistry, Princeton University, Princeton, NJ 08544, USA 18 KEYWORDS. DFT, methane, monooxidation, monooxygenation, iodine 19 20 21 22 ABSTRACT: Recent experiments report high yield (up to 40%) and selectivity (generally > 85%) for the direct partial oxidation of 23 methane to methyl trifluoroacetate in trifluoroacetic acid solvent using hypervalent iodine as the oxidant and in the presence of 24 substoichiometric amounts of chloride anion. We develop here the reaction mechanism for these results based on DFT calculations at the M06-2X/6-311G**++/aug-pVTZ-PP level of plausible intermediates and transition states. We find a mechanistic process that 25 explains both reactivity as well as selectivity of the system. In this oxy-esterification (OxE) system IO Cl − and/or IOCl − act as 26 2 2 4 key transient intermediates, leading to the generation of the high-energy radicals IO2· and Cl· that mediate methane C–H bond 27 cleavage. These studies suggest new experiments to validate the OxE mechanism. 28 29 INTRODUCTION bond of methane, 104 kcal/mol, is stronger than that of methanol (96 kcal/mol) or methyl chloride (100 kcal/mol), 30 Natural gas is a naturally-occurring hydrocarbon mixture leading to abundant over-oxidized side products.[14] 31 composed chiefly of methane that is generally found in the same 32 places as petroleum.[1] However, methane is a gas with a very In the 1960s, Shilov and coworkers discovered that platinum 33 low boiling point of −161.5°C, making it difficult to liquefy, salts in aqueous HCl solution could activate methane under mild [2] conditions, producing methyl chloride (and consequently 34 and consequently expensive to store and transport. Because methanol by hydrolysis) through an electrophilic two-electron 35 natural gas sells for a much lower price than petroleum on the world market, it is often uneconomical to harvest.[3] And C–H activation mechanism.[15-17] In the ensuing decades, much 36 because it is also a potent greenhouse gas with 72 times the work has been done to improve catalyst stability, find less 37 climate change potential of CO2, release into the atmosphere expensive alternatives to platinum, and replace Pt(IV) with 38 can have significantly deleterious environmental effects.[4] It dioxygen as the stoichiometric terminal oxidant.[18-19] One of the 39 has been estimated that 140 billion cubic meters of natural gas most promising developments is the Catalytica process 40 are flared annually.[5] developed by Periana and coworkers, in which methane in oleum solution (H SO /SO ) is oxidized to methyl bisulfate via 41 Much effort has been expended towards partially oxidizing 2 4 3 diverse late-transition metals such as Pt, Pd, Hg, and Au.[20-24] 42 methane to methanol onsite.[6] As a liquid under standard In all of these cases, the reaction is selective for methyl bisulfate 43 conditions and a common chemical feedstock, methanol is both production, often with negligible over-oxidation. This has been 44 easier to store and transport, and also more valuable than attributed to the presence of the electron withdrawing group – methane.[7] A frequently used industrial process is called steam- 45 OSO H which serves as a protecting group, removing electron methane reformation (SMR). A mixture of methane and steam 3 46 density from the methyl C–H bond and slowing over-oxidation is passed through a nickel catalyst at high temperatures and 47 by electrophilic catalysts.[25-26] Currently, the chief drawback of pressures to form synthesis gas (CO and H ), which is then 48 2 the Catalytica family of methane oxidation systems is the use of recombined in a second catalytic process to form methanol. superacidic media such as oleum, which must be constantly 49 Unfortunately, this process is both highly energy and capital replenished due to H O being produced by the reaction, 50 intensive, and it is generally not feasible to be located near 2 rendering cost-effective product extraction very difficult. More 51 remote extraction sites.[8] Another strategy for methane recent work by Periana and coworkers include DFT studies that 52 functionalization is methane oxychlorination, which involves show that superacidic media can be avoided when using main- 53 the catalyzed reaction of methane with HCl and O to form 2 group elements.[27] In particular, a trifluroacetate complex of 54 methyl chloride and water. The products can further react to Sb(V) has been shown to activate methane and ethane.[28] There produce methanol and regenerate HCl.[9-13] Unfortunately, this 55 is a variety of other recent literature showing methane partial process produces a large amount of over-oxidized byproducts 56 oxidation to methanol, syngas, or other products.[29-43] 57 such as methylene chloride, chloroform, and carbon Our work on the partial oxidation of methane has focused on 58 tetrachloride. In fact, a common limitation of radical based mechanisms such as oxychlorination is the fact that the C–H the search for processes efficient under less acidic 59 60 ACS Paragon Plus Environment The Journal of Physical Chemistry Page 2 of 12

environments, in particular trifluoroacetic acid (TFAH).[44] For using a dielectric constant of 8.55 for TFAH.[71] Using the above 1 instance, we have characterized a number of rhodium formula, the probe radius was thereby set to 2.451 Å. 2 complexes with C–H activation ability in TFAH.[45-49] Recently, 3 we have reported the ability of hypervalent iodine complexes to Construction of the free energy 4 oxidize light alkanes (methane, ethane and propane) under mild The state energies for each species in the gas phase were 5 conditions. Hypervalent iodine complexes were reported to [50- calculated using the formulae Ugas = Egas + ZPE + Ulib, Hgas = 6 activate C–H bonds in an electrophilic 2-electron oxidation. 54] Since then, we have also seen that simple iodate and other Ugas + pV, and Ggas = Hgas – TS, where Ulib is the total 7 iodine oxide salts oxidize methane and other light alkanes in the contribution of temperature-dependent librational 8 presence of catalytic amounts of chloride anion.[55-56] In addition (translational, rotational, and vibrational) effects to the internal 9 to 2-election pathways, our work on the hypervalent iodine energy. For each species in the solution phase, the free energy 10 oxide reactions indicate that 1-electron radical pathways should was calculated as Gsol = Ggas + ΔEsolv + ΔGgas→solv, where ΔEsolv 11 be explored. is the solvation energy as described above and ΔGgas→solv = kTln(24.5) represents the free energy change of compressing 1 12 In this contribution, we use density functional theory (DFT) to 13 mol of an ideal gas (volume 24.5 L at 25°C) to 1 L (for 1 M determine the mechanism for the oxidation of methane to standard concentration). In the specific case of individual atoms 14 methyl trifluoroacetate in TFAH solution using hypervalent – or ions such as Br· or Cl , ZPE = 0 and Htot = 5/2kT, and S is 15 iodine oxide as the oxidant. Our results explain current taken from Chase.[72] 16 observations for a range of iodine based oxidants. In particular, 17 we provide a comprehensive characterization of the mechanism Experimental methods 18 of iodate oxidation, and suggest how it may be extended to other 19 iodine sources. In addition, our calculations are valid for TFAH All reactions were prepared in air. Glacial acetic acid (HOAc), and predict non-reactivity in acetic acid. We also present our 20 acetic acid-d4, methyl acetate (MeOAc), ammonium iodate experimental work conducted in parallel to our calculations that (NH IO ), and potassium chloride (KCl) were purchased from 21 4 3 accounts for reactivity in acetic acid and supports our calculated commercial sources and used as received. Methane and argon 22 results.[57] were purchased from GTS-Welco and used as received. 1H 23 NMR spectra were recorded on a Varian Inova 500 MHz NMR 24 MATERIALS AND METHODS spectrometer, using neat samples each with C6D6 in a sealed 25 General calculations capillary as an internal lock reference. Chemical shifts are 1 26 reported relative to standards of nitromethane (CH3NO2) ( H 27 All quantum mechanical calculations were carried out using the NMR δ 4.30). GC-MS data were obtained on a Shimadzu GC- 28 Jaguar software version 8.4 developed by Schrödinger Inc.[58] 2010 instrument equipped with a Restek RT-Qbond 30 m × 8 29 Geometry optimizations were carried out on initial guess mm fused silica PLOT column. Reactions were carried out in 30 structures, and vibrational frequencies were calculated to in-house-built high-pressure stainless steel VCO reactors using Swagelok parts. Each reactor had a Teflon liner and stir bar and 31 confirm the optimized geometries as intermediates (no negative curvatures) or transition states (one negative curvature) and to was heated in an aluminum block. 32 calculate the zero-point energy, entropy, and temperature 33 corrections to obtain the free energy profile. Calculations were RESULTS 34 performed using the M06-2X density functional[59] with the 35 Grimme post-SCF D3 correction for van der Waals The ability of simple iodate salts to oxidize methane was 36 interactions.[60] The double-ζ 6-31G** basis set[61-62] was used reported by Fortman et al.[55] A wide variety of iodine 37 for all elements except bromine and iodine (which used the containing species were investigated in that contribution, and 38 Peterson pseudopotential along with a diffuse-augmented the chemistry was subsequently extended to simple periodate [56] 39 correlation-consistent triple-ζ basis set without f functions, i.e. salts in a follow-up contribution. Table 1 is a summary of the [63-65] 40 aug-cc-PVTZ-PP(-f) ) for geometry optimization and reactions methane undergoes in TFAH, when various iodine oxidants and other additives are added. 41 frequency calculations. All calculations were performed using a temperature set to 498.15 K. A comparison of relevant 42 Table 1. Summary of experimentally determined methane calculated and experimental reaction enthalpies is given in oxidation reactions in TFAH with various iodine oxidants 43 Table S4. 44 and other additives.a 45 Construction of the solvation energy 46 Entry Species Additive Yield MeTFA; MeX 47 Solvation energies ΔEsolv were calculated using the PBF [66-67] 48 Poisson-Boltzmann implicit continuum solvation model using the triple-ζ 6-311G**++ basis set[68-69] for all elements 49 1 I2 − − except bromine and iodine, which used the aug-cc-PVTZ-PP 50 basis set with Peterson pseudopotential and f functions.[63-65] 2 I KCl − 51 This model treats any arbitrary solvent with two parameters: its 2 52 dielectric constant ε and its probe radius r. The probe radius is 53 ퟑ ퟑ횫퐌퐖 calculated according to the formula 풓 = ퟒ훑퐍퐀훒, 3 I2 NH4IO3 2% 54 where r is the probe radius in Å, MW is the molecular weight, 55 NA is Avogadro’s number, ρ is the density, and Δ is the liquid 4 ICl − − 56 packing density and estimated at 0.5 (for comparison, Δ = 0.74 57 for FCC and HCP crystals).[70] Solvent energies were calculated 58 59 60 ACS Paragon Plus Environment Page 3 of 12 The Journal of Physical Chemistry

ΔH = 97.4; ΔG = (4) I + TFAH I + + HI + TFA ― 1 5 ICl3 − 5%; 43% MeCl 2 86.2 2 ΔH = 38.8; ΔG = (5) I + TFAH HI + TFAI 3 2 38.7 6 I(TFA)3 − 7% ΔH = 52.1; ΔG = (6) 4 TFAI TFA ∙ +I ∙ 5 39.5 7 I(TFA) KCl 43% 6 3 Entry 4. ICl has a BDE of 51 kcal/mol, and breaking the bond 7 8 NH IO KCl 21%; <1% MeCl is too endergonic to occur to any significant extent under 8 4 3 reaction conditions (equation 7). However, if we assume that it 9 could break into chlorine and iodine radicals, we see from 10 9 (IO2)2S2O7 − <1% equation 8 that the chlorine radicals are able to break the C–H 11 bonds of methane. The resulting methyl radicals would react with additional ICl to form iodine radicals (equation 9), which 12 10 (IO2)2S2O7 KCl 48% 13 may be able to regenerate additional chlorine radicals 14 (equations 10). Hence our DFT shows that ICl cannot initiate 11 I O − 2%; 15% MeI 15 2 4 methane oxidation, although it may be able to sustain a reaction that has already been initiated by a different radical source. 16 12 I O KCl 30%; 17% MeI a 17 2 4 ICl I ∙ +Cl ∙ ΔH = 51; ΔG = 45.1 (7) 18 ΔHa = −0.2; ΔG = −5.6 (8) Cl ∙ +CH4 HCl + CH3 ∙ 19 13 (IO)2SO4 − 5.3% ΔHa = −30.0; ΔG = −29.5 (9) 20 CH3 ∙ +ICl CH3Cl + I ∙ ΔHa = 14.5; ΔG = 12.9 (10) 21 14 (IO)2SO4 KCl 31% I ∙ +ICl I2 + Cl ∙ 22 aValues based on experimental BDEs.[73] All numbers in kcal/mol. 23 15 KIO4 KCl 41%; 1% MeCl 24 Entry 5. ICl3 actually exists in the solid form as the dimer [78] 25 I2Cl6, although under reaction conditions it is monomeric a 26 Yields for all entries except for 8 and 15 are reported relative (equation 11). It is known to be a powerful oxidizer and chlorine source. In our mechanism, ICl generates chlorine atoms 27 to the amount of iodine source, which was 0.4 mmol. Yields for 3 entries 8 (NH4IO3) and 15 (KIO4) are relative to methane (8.4 (equation 12), which abstract the hydrogen from methane to 28 mmol for NH IO and 2.9 mmol for KIO ). The experiment with generate methyl radicals. The methyl radical abstracts chlorine 29 4 3 4 KIO4 (entry 15) was performed at 200 °C, whereas all other from another equivalent of ICl3 to generate the product, methyl 30 experiments were performed at 180 °C. chloride (equation 13). Finally, radical propagation is achieved 31 through the decomposition of ICl2·, which generates additional 32 In the following sections, we will address the individual chlorine radicals (equation 14). 33 observations in each entry of Table 1. ΔH = 6.1; ΔG = −0.6 (11) 34 Entries 1-2. The lack of reaction with elemental iodine is of I2Cl6 2ICl3

35 little surprise, as iodine radicals are generally thought of as ICl ICl ∙ +Cl ∙ ΔH = 41.0; ΔG = 28.3 (12) 36 3 2 being too weak to break the C–H bonds of methane. This is ΔHa = −0.2; ΔG = (8) 37 confirmed by our calculations (equations 1-3). From equation Cl ∙ +CH4 HCl + CH3 ∙ −5.6 38 2, we see that abstracting a hydrogen atom from methane by an ΔH = −46.9; ΔG = (13) 39 iodine radical is endergonic by 25.8 kcal/mol. More CH3 ∙ + ICl3 CH3Cl + ICl2 ∙ −46.3 40 importantly, by combining equations 2 and 3 we see that ΔH = 25.7; ΔG = 17.9 (14) conversion of methane and iodine to methyl iodide and ICl2 ∙ ICl + Cl ∙ 41 a [73] hydrogen iodide is both endothermic and endergonic. Values based on experimental BDEs. All numbers in kcal/mol. 42 43 ΔHa = 36.5; ΔG = 32.2 (1) This is our first entry in which methane is oxidized to a 44 I2 2I ∙ significant extent, although the main product is methyl chloride. ΔHa = 31.7; ΔG = 25.8 (2) 45 I ∙ +CH4 HI + CH3 ∙ Equations 11 and 12 show that dissociation to the monomer ICl3 46 ΔHa = −19.0; ΔG = −16.9 (3) is favorable and I–Cl bond homolysis is accessible. According CH3 ∙ + I2 CH3I + I ∙ to experimental observations,[55] methyl chloride does not 47 a [73] Values based on experimental BDEs. All numbers in kcal/mol. convert to methyl trifluoroacetate under reaction conditions. 48 However, methyl iodide is rapidly converted via S 2 attack to Entry 3. Although I and KIO have been reported to oxidize N 49 2 3 methyl trifluoroacetate, even at room temperature, but only 50 methane in oleum,[74-77] the same reaction was not observed in under basic conditions such as for the NH4IO3/KCl case (entry 51 the more weakly acidic TFAH. The oleum reaction is proposed + 8) or the KIO4/KCl case (entry 17). 52 to involve I as a reactive intermediate, but we find that this species is too unstable to be accessible in TFAH (equation 4). 53 Entries 6-7. In 2014, Konnick et al. reported the ability of the TFAI may be seen as a lower-energy analogue (equation 5); 54 hypervalent I(III) compound (C6F5)I(TFA)2 to oxidize methane however, the O–I bond dissociation energy is still too high to MeTFA in TFAH solution.[79] Their mechanism, supported 55 (equation 6). 56 by DFT calculations, involved the electrophilic 2-electron activation of methane to form an organoiodine(III) 57 58 59 60 ACS Paragon Plus Environment The Journal of Physical Chemistry Page 4 of 12

intermediate, which is then attacked via SN2 to form MeTFA. In comparing entries 6 to 7, we see that the addition of chloride 1 However, our calculations which examine I(TFA)3 instead find greatly improves yield from 7% to 43%. We first consider the − 2 that the analogous closed-shell transition state is too high in possibility of I(TFA)3 and Cl forming adducts (equations 16- 3 energy (Scheme 1). Hence, we believe an alternative 18), and we see that substitution of TFA by Cl is favorable for 4 mechanism is needed to explain the reactivity in entries 6 and all three reactions. However, we consider that only the 5 7. formation of I(TFA)2Cl is feasible since only substoichiometric 6 amounts of KCl are added. Scheme 1. Closed shell mechanism of methane oxidation by 7 ― ΔH = −6.5;― ΔG = −14.5 (16) I(TFA)3. I(TFA)3 + Cl I(TFA)2Cl + TFA 8 ― ΔH = −2.9;― ΔG = −11.2 (17) 9 I(TFA)2Cl + Cl I(TFA)Cl2 + TFA CF3 CF3 CF3 ― ΔH― = −6.8; ΔG = −12.2 (18) 10 O O O I(TFA)Cl2 + Cl ICl3 + TFA TFA TFA TFA 11 I O I O I O TFA H = 22.9 TFA H = -60.7 TFA From I(TFA)2Cl, we may expect thermal homolysis to break 12 Me H Me H Me H 13 G = 41.3 G = -85.1 either the I–TFA (equation 19) or I–Cl (equation 20) bond. Either way, radicals are formed that can activate methane as in 14 H = 14.5 G = 34.7 Scheme 2. 15 16 TFA TFA I(TFA)2Cl ΔH = 48.8; ΔG = (19) 17 I Me O I Me O 27.7 CF3 CF3 I(TFA)Cl ∙ +TFA ∙ O O 18 H = -22.6 ΔH = 47.8; ΔG = (20) O O I(TFA) Cl I(TFA) ∙ +Cl ∙ 19 H G = -56.7 H 2 2 34.8 20 F3C O F3C O 21 The resulting methyl radicals can react with additional 22 Instead, we propose a radical mechanism initiated by the I(TFA)2Cl to form MeTFA (equation 21) in preference to MeCl homolysis of an I–TFA bond (Scheme 2). Both radical (equation 22), which provides an explanation for why only 23 fragments TFA· and I(TFA)2· can then either activate methane MeTFA is observed as an oxidized product. 24 (red) or decompose (blue). Finally, the methyl radical can react 25 ΔH = (21) with another equivalent of I(TFA)3 to form the product MeTFA 26 −52.8; and propagate the reaction (equation 15). CH3 ∙ + I(TFA)2Cl CH3TFA + I(TFA)Cl ∙ 27 ΔG = 28 Scheme 2. Open shell mechanism of methane activation by −55.1 I(TFA) a. ΔH = (22) 29 3 −40.0; 30 TFA CH3 ∙ + I(TFA)2Cl CH3Cl + I(TFA)2 ∙ H = 20.7 ΔG = I H 31 G = 33.4 −39.8 32 CH4 TFA -CH3 33 H = 25.7 Non-radical paths for CH activation using potential I(III) I 4 34 TFA TFA TFAI G = 9.2 intermediates were sought, but no thermally accessible TFA transition states were found (see Supporting Information, 35 I TFA O Scheme S1). 36 TFA H = 48.8 CH4 H = -18.6 G = 22.1 TFAH 37 F3C O G = -20.2 -CH3 Entry 8. The reaction of methane and iodate salts in the 38 H = 13.0 TFAH H = -23.8 presence of catalytic amounts of chloride is the chief focus of 39 -CO2 [55] G = 13.0 -CF3H G = -37.0 the report by Fortman et al. and may be considered the 40 b CH4 H = -2 “canonical” hypervalent iodine oxidation reaction. Hence, we CF3 F3CH 41 -CH3 G = -7.3 devote the bulk of our analysis to this reaction and its variations. 42 a b We begin by noting that while iodate is quite basic in TFAH 43 Methane activation in red; decomposition in blue. Values based on experimental BDEs.[73] All numbers in kcal/mol. solution under reaction conditions (equation 23) and thus 44 protonated to HIO3, chloride, iodide, and water are all weak 45 ΔH = −52.9; (15) bases (equations 24-26, respectively) and remain unprotonated. 46 CH3 ∙ + I(TFA)3 CH3TFA + ITFA2 ∙ ΔG = −54.8

47 ― ―ΔH = 0.3; ΔG = −2.4 (23) IO3 + TFAH HIO3 + TFA 48 From a close examination of Scheme 2, we see that a number ― ― ΔH = 10.6; ΔG = 6.9 (24) of decomposition pathways may affect the yield of CH3· radical Cl + TFAH HCl + TFA 49 ― ― ΔH = 19.3; ΔG = 15.1 (25) 50 generation, hence explaining the low yield observed for entry 6. I + TFAH HI + TFA Our case is bolstered by the experimental observation of 51 H O + TFAH H O + + TFAΔ―H = 26.4; ΔG = 25.2 (26) fluoroform formation as a side product as reported by Konnick 2 3 52 et al.,[79] indicating that their hypervalent iodine reaction may Furthermore, we note that if elemental iodine is in the system 53 have a radical component as well. Also, evidence of radical CF 3 (as it is the eventual fate of iodate in the reaction[55-56]), it can 54 formation is found for the conversion of methane and complex to chloride and iodide (equations 27-28, respectively). 55 iodate/chloride in TFAH.[55-56] 56 This would further decrease the thermodynamic activity of these halides and thus make their protonation even less likely. 57 58 59 60 ACS Paragon Plus Environment Page 5 of 12 The Journal of Physical Chemistry

− Cl ― + I I Cl ― ΔH = −10.9; ΔG = −3.4 (27) Both ICl and IOCl3 can generate Cl radicals: ICl through 2 2 − 1 reaction with I·, and IOCl3 through auto-dissociation ― ― ΔH = −9.4; ΔG = −2.9 (28) 2 I + I2 I3 (equations 31-32). 3 a Having established that iodate is initially protonated to HIO3, ΔH = 14.5; ΔG = 12.9 (10) 4 I ∙ +ICl I2 + Cl ∙ we wish to ascertain if the TFAH environment may result in 5 ― ― ΔH = 26.6; ΔG = 17.7 (31) TFA/aqua/chloro adducts being formed. In order to approach IOCl3 ∙ IOCl2 + Cl ∙ 6 a [73] this rigorously, we begin by noting that HIO3 is an I(V) complex Values based on experimental BDEs. All numbers in kcal/mol. 7 with only oxo and hydroxo ligands. In water, an oxo ligand can 8 interconvert with two hydroxo ligands. In TFAH, an oxo can Both IO2· and Cl· radicals are able to initiate hydrogen 9 interconvert with two TFA ligands. And in the presence of a abstraction from methane (equations 32 and 8, respectively). 10 chloride salt, hydroxo and TFA ligands may interconvert with The HCl produced by equation 8 is then deprotonated and absorbed by I (equations 24 and 27), while the HIO produced 11 chloro ligands. Thus, we would expect HIO3 in wet TFAH with 2 2 12 chloride to exist as a multitude of iodine complexes with a by equation 32 disproportionates to HIO3 and HOI (equation 33), which can then convert to additional ICl (equation 34). A 13 combination of oxo, hydroxo, chloro, and TFA substituents such that the overall oxidation state of the iodine is +5. more detailed treatment of HIO2 and other I(III) species can be 14 found below in the section for entries 11-14. 15 A complete accounting of I(V) species in TFAH/H2O with 16 optional chloro ligands, along with their relative energies, is ΔH = (32) 17 given in Table S1. From this table, we can extract the lowest 14.3; 18 energy I(V) species containing at least one I–Cl bond (Scheme IO2 ∙ +CH4 HIO2 + CH3 ∙ ΔG = 19 S2). By calculating the difference in energy between each 10.3 20 species and the corresponding species with the I–Cl bond ΔHa = (8) broken, we can find the BDEs and free energies for 21 −0.2; homolytically cleaving each I–Cl bond (see Supporting Cl ∙ +CH4 HCl + CH3 ∙ ΔG = 22 Information, Scheme S2, in red). These values are in excellent −5.6 23 agreement with the previous coupled-cluster work by Dixon ΔH = (33) [80] 24 and coworkers. However, we did not find a species with a −21.6; 25 readily homolyzible bond. The lowest energy species are 2HIO2 HIO + HIO3 ΔG = – – 26 IO2Cl2 and IOCl4 , but breaking the I–Cl bond of these species −21.3 27 is endergonic by 44.1 and 37.7 kcal/mol, respectively, which is ΔH = (34) 28 too unfavorable to be considered as contributing processes. By 0.4; HIO + I Cl ― + TFAH ICl + I + H O + TFA ― 29 contrast, the species IOCl3 and ICl5 can have their I−Cl bonds 2 2 2 ΔG = broken at a free energy cost of 23.8 and 17.1 kcal/mol, −10.2 30 − respectively, but they are significantly less stable than IO2Cl2 aValues based on experimental BDEs.[73] All numbers in kcal/mol. 31 − (Grel of 10.3 and 20.1 compared to −6.5 for IO2Cl2 ) and hence 32 not likely to be present in solution. Due to the presence of elemental iodine in the reaction, the 33 methyl radicals produced will most likely react with I2 to 34 In considering how the reaction might proceed, we decided to produce CH3I (equation 35), which is converted to the product 35 investigate more closely the role of elemental iodine. We have CH3TFA under reaction conditions (equation 36). Finally, the 36 observed experimentally that elemental iodine is formed as a iodine radicals can recombine to form additional I2 (equation 37 final product in our reaction.[55] Elemental iodine is well known 37), and the iodide produced in equation 36 can 38 to be a radical inhibitor, due to the relative inertness (as far as comproportionate with HIO to produce additional elemental [81] 39 radicals go) of I·. Hence, we decided to add the assumption iodine (equation 38). that the system at steady state would include a significant 40 ΔHa = −19.0; ΔG = −16.9 (35) concentration of I·, either produced from 1), the reaction of CH3 ∙ + I2 CH3I + I ∙ 41 other radicals with I2, or 2), the autolysis of I2 itself, which CH I + TFA ― CH TFA +ΔHI ― = −9.4; ΔG = −2.9 (36) 42 according to equation 1 is endergonic by a relatively accessible 3 3 ΔHa = −36.5; ΔG = −32.2 (37) 43 32.2 kcal/mol. A steady state approximation for the reaction 2I ∙ I2

44 kinetics should also affect our calculation for the overall ― ΔH = −21.0;― ΔG = −23.7 (38) I + HIO + TFAH I2 + H2O + TFA 45 reaction barrier, given later in this section. aValues based on experimental BDEs.[73] All numbers in kcal/mol. 46 47 We thus look at the reaction of several I(V) species with I· Scheme 3 shows the synthesis of the above equations into a full 48 (Figure S1), and we are gratified to find that the abstraction of mechanism for the conversion of methane to methyl Cl is generally facile. In particular, we see that our low energy trifluoroacetate for the iodate-based OxE process. The 49 − − anionic species IO2Cl2 and IOCl4 also have the lowest barriers mechanism proceeds through typical phases of initiation, 50 − for reacting with I·, generating ICl and IO2Cl (which propagation, and termination. During initiation, the lowest 51 − − spontaneously decomposes to IO2· and Cl ) or IOCl3 , energy I(V)−Cl species are formed, and I· radical is generated. 52 respectively (equations 29-30). Propagation begins with the I· radical reacting with the I(V) 53 species to create the more reactive radicals IO2· and Cl· 54 ― ― ΔG‡ = 4.5; ΔG = 0.6 (29) I ∙ +IO2Cl2 ICl + IO2Cl ∙ (reactive radical activation). The radicals IO2· and Cl· then

55 ― ― abstract hydrogen atoms from methane, which rebounds on IO2Cl ∙ IO2 ∙ + Cl 56 elemental iodine to form methyl iodide and regenerate I· ― ― ΔG‡ = 8.3; ΔG = –7.4 (30) 57 I ∙ + IOCl4 ICl + IOCl3 ∙ (methane activation and iodine radical regeneration). Finally, 58 59 60 ACS Paragon Plus Environment The Journal of Physical Chemistry Page 6 of 12

methyl iodide can undergo S 2 attack by TFA− to produce the 1/2 ― N [I2] [IO2Cl2 ]. Hence, the rate constant is 푘 = 푘29 푘1/푘 ―1. 1 methyl TFA product, and the iodide and HIO can ‡ 2 푘퐵푇 ∆퐺 [83-84] 2 disproportionate and comproportionate to form additional Using the Eyring-Polanyi equation 푘 = ℎ exp ( ― 푅푇 ), iodate and elemental iodine (iodate regeneration). 3 we see that overall activation barrier is then ∆퐺 ‡ = ―푅푇ln 4 ℎ ‡ 1 ‡ 1 ‡ In order to estimate the overall kinetic barrier for Scheme 3, we 푘29 푘1/푘 ―1 = ∆퐺 ― 푅푇ln(푘1/푘 ―1) = ∆퐺 + ∆퐺 5 (푘퐵푇 ) 29 2 29 2 1 consider a steady-state approximation for [I·], using the method 6 1 ‡ in Carey and Sundberg’s Advanced Organic Chemistry page ― 2∆퐺 ―1. If we assume that the dissociation and reformation 7 [82] ‡ 683. We approximate that the rate of formation and of I has minimal barrier, we can approximate ∆퐺 ≈ ∆퐺1 and 8 2 2 1 destruction of I· to be equal, i.e. 푘1[I2] = 푘 ―1[I ∙ ] , where 푘1 ‡ ‡ 9 2 ∆퐺 ―1 ≈ 0. Since we have calculated ∆퐺29 as 4.5 kcal/mol [I2] is the rate of initiation (equation 1) and 푘 ―1[I ∙ ] the rate of 10 1/2 (Scheme 3) and ∆퐺1 as 32.2 kcal/mol (equation 1), we estimate [I ∙ ] = 푘1/푘 ―1[I ] 11 termination. Then, 2 . We can also say that the effective barrier to be about 20.6 kcal/mol. the rate of the reaction is the rate of a propagation step, for 12 ― 푘 /푘 13 example equation 29: rate = 푘29[I ∙ ][IO2Cl2 ] = 푘29 1 ―1 14 15 Scheme 3. Overall summary of the mechanism for the iodate-based OxE process for methane oxidation. All numbers in 16 kcal/mol. *Values based on experimental BDEs.[73] 17 18 Initiation: 19 - - - I 2I H = 36.5*, G = 32.2 20 HIO3 + 2I2Cl + TFAH IO2Cl2 + 2I2 + H2O + TFA H = 3.0, G = -6.5 2 21 HIO + 4I Cl- + 3TFAH IOCl - + 4I + 2H O + 3TFA- H = 14.2, G = -6.1 22 3 2 4 2 2 23 Propagation: 24 25 Reactive radical activation H1 = -9.6, G1 = -0.8 26 - I Cl I Cl - H = -2.0, G = 4.5 1 2 I + IO2Cl2 ICl + Cl + IO2 I + ICl I I Cl I2 + Cl H = 22.3, G = 13.7 O O H = 12.5, G = 0.6 2 2 27 H = 14.5*, G = 12.9 28 O O 29 Cl Cl Cl Cl Cl Cl Cl Cl H = -0.7, G = 8.3 Cl I Cl I + I I ICl + I I Cl + H = 26.6, G = 17.7 30 H = -2.1, G = -7.4 Cl Cl Cl Cl O Cl O Cl O 31 I 32 33 Methane activation 34 I O 35 IO2 + CH4 H3C H O CH3 + HIO2 H = 14.3*, G = 10.3 Cl + CH4 H3C H Cl CH3 + HCl H = -0.2*, G = -5.6 36 Iodine radical and iodate regeneration 37

38 CH3 + I2 H3C I I CH3I + I H = -19.0, G = -16.9 2HIO2 HOI + HIO3 H = -21.6, G = -21.3 39 40 Termination and product formation 41 2I I H = -36.5*, G = -32.2 CH I + TFA- CH TFA + I- H = -9.4, G = -2.9 42 2 3 3 - - - - 43 HOI + I2Cl + TFAH ICl + I2 + H2O + TFA H = 0.4, G = -10.2 HOI + I + TFAH I2 + H2O + TFA H = -21.0, G = -23.7 44 45 Equation 39 shows the overall reaction of the iodate/chloride reactions in the absence of methane gave statistically identical 46 oxidation. MeOAc yields as reactions that include methane, approximately 47 30% yield with respect to oxidant using 0.67 mmol KCl and 7.7 ΔH = (39) 48 mmol of NH4IO3 in acetic acid. Furthermore, GC-MS analysis −234.1; of the reaction of perprotio-methane in acetic acid-d4 indicated 49 2HIO3 + 5CH4 + 5TFAH I2 + 6H2O + 5CH3TFA ΔG = that the fully deuterated derivative CD3CO2CD3 accounts for > 50 −228.3 90% of methyl acetate production, while the product that would 51 arise from methane functionalization, CD3CO2CH3, accounts 52 Entry 8, acetic acid variation. Previously, we showed that the for < 2% of the methyl acetate product. 53 reaction of methane and iodate/chloride in acetic acid produces [55] Using DFT, it is readily apparent that the same mechanism 54 methyl acetate (MeOAc). However, on further evaluation, it was found that MeOAc formed under these conditions is the proposed for the iodate-based OxE process in TFAH when 55 translated to acetic acid solvent will not work. The biggest 56 result of the reaction with acetic acid only and that the majority of the MeOAc is not derived from methane oxidation. Control difference between acetic acid and TFAH is the much milder 57 acidity of acetic acid. In equation 23, we see that iodate is quite 58 59 60 ACS Paragon Plus Environment Page 7 of 12 The Journal of Physical Chemistry

basic in TFAH and is expected to be fully protonated. However, ― + ΔH = 10.3; (50) 1 by comparison we can see from equation 40 that iodate is a H2S2O7 + H2O HS2O7 + H3O ΔG = 8.3 2 weak base in HOAc and is expected to remain in the anionic ΔH = 23.1; (51) H SO + H O HSO ― + H O + 3 form. Chloride is expected to remain anionic in both cases 2 4 2 4 3 ΔG = 21.9

4 (equations 24 and 41); however, in HOAc it is not expected to ― + ΔH = 15.7; (52) HCl + H2O Cl + H3O 5 be absorbed by elemental iodine (compare equation 42 to 27). ΔG = 18.2 ΔH = 7.1; ΔG (53) 6 ― + a HI + H2O I + H3O ― ― ΔH = 17.0; ΔG = (40) = 10.0 7 IO3 + HOAc HIO3 + OAc 13.7 8 a ΔH = 32.3; ΔG = (41) We also note that with the conversion of (IO2)2S2O7 to HIO3 and Cl ― + HOAc HCl + OAc ― 9 27.9 H2SO4, the system is still not acidic enough to remove chloride 10 a ― ― ΔH = −6.2; ΔG = (42) from its sequestration by elemental iodine (equations 54-55). 11 Cl + I2 I2Cl 1.3 ΔH = 5.5; (54) 12 aValues calculated for HOAc solution All numbers in kcal/mol. ― ― H2S2O7 + I2Cl HS2O7 + I2 + HCl ΔG = −6.6 13 ΔH = 18.3; (55) The calculated energies of various I(V)−Cl species compared to H SO + I Cl ― HSO ― + I + HCl 14 − − 2 4 2 4 2 ΔG = 7.0 this new IO3 /Cl baseline reveal that they are much higher in 15 energy (Figure S2). We are simply unable to form the same 16 quantities of the I(V)−Cl intermediates necessary to generate Although in basic solution the energies of I(V) species can be − − 17 high energy radicals. We also note that the H−O bond in acetic referenced to the species HIO3, I2Cl , I2, H2O, TFAH, and TFA , 18 acid is much weaker (BDE 112 kcal/mol) than that in TFAH in acidic solution they should be referenced to the species HIO3, − − 19 (124.5 kcal/mol); hence, radicals that are generated are much I2Cl , I2, H2O, TFAH, H2SO4, and HSO4 . The columns in Table 20 more likely to react with the acetic acid solvent, abstracting H S1 that are marked “acid” show the energies of the I(V) species atoms to form species such as ·CH CO H and CH CO ·, which recalculated to this new reference. As in the basic case of the 21 2 2 3 2 − − may decompose to liberate CO and form other radicals such as IO3 /Cl system, the lowest energy species formed are still 22 2 − − − CH · that ultimately become MeOAc. IO2Cl2 and IOCl4 , with the only difference being that IOCl4 23 3 − is now favored over IO2Cl2 . Scheme S3 shows the new 24 − − initiation steps for the methane oxidation in acidic condition Entries 9-10. The mechanism for IO3 and Cl as described for 25 entry 8 can easily hold if the iodate source is replaced by (i.e., (IO2)2S2O7), whereas the propagation and termination 26 steps remain the same as for Scheme 3. (IO2)2S2O7. We would expect any water in the system to 27 hydrolyze (IO2)2S2O7 to HIO3 and H2SO4 (equations 43-44). 28 Hence, it is expected the same mechanism with improved yield Entries 11-14. I2O4 and (IO)2SO4 can be analyzed as ionic n+ 29 due perhaps to the improved acidity of the system. species, consisting of a linear polymeric cationic (IO)n chain with either IO − or SO 2− as the counterion.[85-86] The iodine 30 3 4 ΔH = (43) atoms in (IO) n+ have a formal positive charge, and the 31 n (IO2)2S2O7 + 2H2O 2HIO3 + H2S2O7 −59.1; ΔG counterions may be coordinated to them (strongly so in the solid 32 = −65.6 and in low dielectric solvents; more weakly in high dielectric 33 ΔH = (44) [86] solvents such as ethanol and dimethylformamide). We see n+ 34 H2S2O7 + H2O 2H2SO4 −26.8; ΔG that the (IO)n chain itself must be active in effecting methane 35 = −32.2 oxidation, since entry 14 shows that the reaction works with an 36 inert sulfate counterion. In general, it is known that I(V) salts Note that HS O − and HSO −, like Cl− and I−, are weak bases in 37 2 7 4 are more stable than I(III), and indeed, both I2O4 and (IO)2SO4 TFAH, but S O 2− and SO 2− are favorable in reacting with 38 2 7 4 are known to disproportionate to HIO3 and I2 in hot water and TFAH and are thus protonated (equations 45-48). [85] 39 sulfuric acid, respectively. We would thus expect HIO3 to be likely formed in hot TFAH as well. 40 ― ― ΔH = 16.0; (45) HS2O7 + TFAH H2S2O7 + TFA n+ 41 ΔG = 16.9 Our computational studies initially indicated that (IO)n chains + 42 2 ― ― ― ΔH = −4.6; (46) in TFAH are unstable with respect to dissociation into I=O S2O7 + TFAH HS2O7 + TFA 43 ΔG = −3.9 monomers (equation 56).[87-88] However, iodate or sulfate is the ΔH = 3.3; ΔG (47) counterion, and both are very basic (equations 23 and 48), 44 HSO ― + TFAH H SO + TFA ― 4 2 4 = 3.3 leading to the formation of TFA−, which can very favorably 45 + 2 ― ― ― ΔH = −24.0; (48) react with IO to form TFA−I=O (equation 57). The TFA−I=O 46 SO4 + TFAH HSO4 + TFA ΔG = −26.8 produced can further add a water or TFAH molecule across the 47 I=O double bond (equations 58-59), or polymerize (equations + 48 Also note that H3O is a weak acid in TFAH (equation 49), and 60-61). Table S2 shows a comprehensive listing of all 49 since the reaction generates H2O as a byproduct, we would monomeric I(III) species that may be formed in TFAH solution. expect H O+ and not TFAH + to be the active Brønsted acid in 50 3 2 The lowest energy monomeric species is ICl3 in both basic and acidic wet TFAH solution. Although H S O , H SO , HCl, and 51 2 2 7 2 4 acidic systems. Since the polymerization of I(OH)TFA2 is HI are all considered strong acids in aqueous solution, due to 52 approximately thermoneutral (equation 61), and since ICl3 is the much lower dielectric constant of TFAH, none of them are 53 significantly lower in energy than I(OH)TFA2 (equation 62), we likely to protonate H O in TFAH solution (equations 50-53). 54 2 can restrict our mechanistic considerations to monomeric I(III) 55 ΔH = 20.3; (49) species only. H O + + TFAH H O + TFAH + 56 3 2 2 ΔG = 21.5 57 58 59 60 ACS Paragon Plus Environment The Journal of Physical Chemistry Page 8 of 12

ΔH ~ −17 to (56) ΔH = −35.6; (69) ― ― 푛 + + (푛 ― 1) + [88] HOCl + HIO2 + I2 + TFA I2Cl + HIO3 + TFAH 1 (IO)푛 IO + (IO)푛 ― 1 −20; ΔG ~ −30 ΔG = −22.8 ― 2 to −33 HOCl + ICl + H2O + 2I2 + 2TFA ΔH = −14.5; (70)

3 ΔH = −39.1; ΔG (57) ― ΔG = 8.7 IO + + TFA ― I(O)TFA 2I2Cl + HIO2 + 2TFAH 4 = −25.3 5 ΔH = −24.4; ΔG (58) − − I(O)TFA + H2O I(OH)2TFA Because IO3 is generated from IO4 (equation 66), our = −9.2 − − 6 mechanism for IO3 (entry 8) should also be valid for IO4 . ΔH = −24.8; ΔG (59) − 7 However, since IO4 can be expected to make a much richer set I(O)TFA + TFAH I(OH)TFA2 = −6.1 of chloro, TFA, oxo, and hydroxo complexes than IO −, we have 8 ΔH ~ −29 to (60) 3 only demonstrated the feasibility of one possible reaction 9 −24; ΔG ~ −8 to [89] H[I(O)TFA]푛TFA + I(O)TFA H[I(O)TFA]푛 + 1TFA pathway (IO − to methane oxidation via IO −) without 10 −4 4 3 discounting the possibility of other pathways. Equation 71 11 ΔH ~ −4 to 1; (61) shows the overall reaction of the periodate/chloride oxidation, 12 H[I(O)TFA]푛TFA + I(OH)TFA2 HΔ[IG(O ~) TFA−2 to]푛 2+ 1TFA +[90]TFAH ― which is even more downhill than iodate/chloride oxidation 13 I(OH)TFA2 + 3I2Cl + TFAH ΔH = 26.6; ΔG = (62) (compare to equation 39). 14 ― −15.5 ICl3 + 3I2 + H2O + 3TFA 15 ΔH = (71) −374.4; 16 We now see that the species involved here form the same 2HIO4 + 7CH4 + 7TFAH I2 + 8H2O + 7CH3TFA ΔG = 17 collection of species in the reactions of entry 5 (ICl3) and entry − −372.5 18 7 (ITFA3/Cl ). As in the case of entry 7, MeTFA and not MeCl is formed due to the catalytic amounts of Cl− preventing the 19 DISCUSSION 20 build-up of ICl3 itself. We can also consider a 21 disproportionation route to form I(V) that can enter that As mentioned above, a common limitation of radical mechanistic route. In one such example, monomeric I(O)TFA 22 mechanisms is the difficulty of preventing over-oxidation. For favorably disproportionates to IO TFA (equation 63), which 23 2 example, free radical chlorination of methane at reasonable can then convert to HIO3 (equation 64). The I(V) mechanism as 24 conversions often yields a mixture of methyl chloride, for the iodate reaction (entry 8) then applies. The TFAI [91] 25 methylene chloride, chloroform, and/or carbon tetrachloride. byproduct can convert to ICl (equation 65) and enter the I(V) Methane oxidation by electrophilic C–H activation can be 26 mechanism from there. selective for CH3X (X = TFA, SO3H, etc.) due to the electron- 27 [25-26] ΔH = −22.9; ΔG (63) withdrawing nature of X acting as a protecting group. In 28 2I(O)TFA IO2TFA + ITFA = −20.7 the iodate/chloride OxE process, methane and ethane were 29 ΔH = −10.0; ΔG (64) oxidized to CH3TFA and EtTFA respectively, with conversions 30 IO2TFA + H2O HIO3 + TFAH = −12.8 greater than 20% and with formation of minimal over-oxidized [55] 31 products. Is it possible that TFA could serve as a protecting ― ― ΔH = 1.9; ΔG = (65) 32 TFAI + I2Cl ICl +I2 + TFA −10.0 group against radical oxidation? 33 Consider the immediate result of a radical H-atom abstraction Entry 15. Periodate can actually refer to two different ions: 34 on CH TFA: the formation of the radical ·CH TFA. A case metaperiodate (IO −) and the water adduct orthoperiodate 3 2 35 4 could be made that a lone pair on the O atom of TFA directly (H IO −).[85] Whereas orthoperiodate is generally known to be 36 4 6 bound to the CH can stabilize the radical via the formation of the preferred species in aqueous solution, we wanted to 2 37 a two-centered, three-electron bond.[82] This would make ascertain if metaperiodate persists under our reaction conditions radical attack on CH TFA and thus over-oxidation more likely, 38 (potentially wet TFAH). Scheme S4 shows that both addition of 3 similar to observations for methane oxychlorination. A case 39 water and protonation to IO − is unfavorable; hence, we can 4 could also be made that in the transition state [TFACH ---H--- 40 begin our analysis with the IO − species itself. 2 4 Cl]·‡, the C has partial cationic character due to the 41 electronegative nature of the Cl atom. The presence of the In the presence of elemental iodine, comproportionation to 42 electron-withdrawing TFA group thus destabilizes the iodate is favorable (equation 66). In the absence of I , chloride 43 2 transition state in what is termed a radical polar effect, and may serve as a reducing agent instead (equations 67-68). 44 hence inhibits over-oxidation.[92] 45 ― ― ΔH = −27.6; (66) 46 5IO4 + I2 + H2O 5IO3 + 2HIO3 ΔG = −40.4 In order to investigate the competing trends of lone pair 47 ΔH = 24.0; ΔG (67) stabilization and polar effect destabilization, we examine IO ― + Cl ― IO ― + ClO ― 48 4 3 = 18.5 several radical abstraction reactions and compare them to ΔH = −18.9; (68) equation 8 (equations 72-76). Our insights from the 49 ClO ― + TFAH HOCl + TFA ― 50 ΔG = −20.6 comparisons are summarized in Table 2. 51 As shown in Scheme 3, in the presence of chloride the iodate ΔG‡ = 3.2; ΔG = −5.6 (8) 52 Cl ∙ +CH4 HCl + CH3 ∙ can oxidize methane to CH3TFA and generate HIO2 (equation ‡ 53 Br ∙ +CH HBr + CH ∙ ΔG = 15.5; ΔG = 8.6 (72) 32), which then disproportionates to regenerate HIO3 through 4 3 54 ΔG‡ = 6.1; ΔG = −6.3 (73) several other iodine oxyacids (equations 33-34). The Cl ∙ +CH3TFA HCl + TFACH2 ∙ 55 hypochlorous acid produced in equations 67-68 can react with ΔG‡ = 19.6; ΔG = 7.9 (74) Br ∙ +CH3TFA HBr + TFACH2 ∙ 56 these iodine oxyacids, aiding in the regeneration of HIO3 ΔG‡ = 3.6; ΔG = −7.6 (75) 57 (equations 69-70). Cl ∙ +CH3Cl CH2Cl ∙ +HCl 58 59 60 ACS Paragon Plus Environment Page 9 of 12 The Journal of Physical Chemistry

‡ Cl ∙ +CH Cl CHCl ∙ +HCl ΔG = 4.3; ΔG = −8.9 (76) reacts with iodine radicals present in the system to form IO2· 1 2 2 2 Table 2. Comparison of radical H-abstractions to isolate and Cl· radicals. These radicals are able to mediate the C−H 2 polar and lone pair effects. bond cleavage of methane, leading to CH3I and ultimately the 3 final product, CH3TFA, which is protected from further 4 oxidation by the polar effect. ‡ 5 Comparison ΔΔG ΔΔG We thus report a simple and unifying solution to the mechanism 6 of the iodate/OxE process, a problem that initially appears to 7 involve an intractable soup of chemical species. These insights CH4 v. CH3TFA + Cl· (eqs. 8 v. 73) 2.9 −0.7 8 may serve as a starting point for further investigations into this 9 class of oxidation reactions. CH v. CH TFA + Br· (eqs. 72 v. 74) 4.1 −0.7 10 4 3 11 ASSOCIATED CONTENT 12 CH4 v. CH3Cl + Cl· (eqs. 8 v. 75) 0.4 −2.0 Supporting Information. Additional schemes and tables and a 13 spreadsheet containing raw data for all chemical species studied. This material is available free of charge via the Internet at 14 CH Cl v. CH Cl + Cl· (eqs. 75 v. 76) 0.7 −1.3 3 2 2 http://pubs.acs.org. 15 16 As shown in the final column of Table 2, the calculated ΔΔGs AUTHOR INFORMATION 17 confirm the lone pair stabilization effect: In all cases, ΔG for the 18 Corresponding Author reaction Y ∙ + CX H HY + CX H ∙ is lower if X is a 19 푛 4 ― 푛 푛 3 ― 푛 *Author to whom correspondence should be addressed. Email: substituent such as Cl or TFA as opposed to H. 20 [email protected] 21 From examining the ΔΔG‡ column instead, we see that the Author Contributions 22 transition state for H atom abstraction for CH3TFA + Y· (Y = All calculations were carried out by RF, with strategies suggested 23 Cl or Br) is about 3-4 kcal/mol higher than the corresponding by RF, RJN, and WAG. Experimental work was carried out by 24 transition state for CH4 + Y·. This is the polar effect, and it can NSL under mentorship by TBG. Critical guidance was provided 25 be confirmed by Mulliken charges (Table S3). The charge on by JTG and TBG. The manuscript was written through contributions of all authors. All authors have given approval to 26 carbon in CH4 is −0.60 due to the relative electronegativity of the final version of the manuscript. 27 carbon versus hydrogen. For CH3TFA and CH3Cl, this 28 increases to −0.33 and −0.40, respectively; due to the electron Funding Sources withdrawing TFA or Cl groups, which have Mulliken charges 29 This research was sponsored by the U.S. Department of Energy, −0.27 and −0.12, respectively. In the transition states [CH3---H- 30 ‡ ‡ Office of Energy Efficiency and Renewable Energy, Advanced --Cl]· and [CH3---H---Br]· , the charge on C is increased to 31 −0.52 and −0.51, respectively; an illustration of the polar effect Manufacturing Office, under contract DE-AC05-00OR22725 with 32 due to the electronegative Cl and Br atoms. In the comparison UT-Battelle, LLC. The funding for this computational project came from various gifts to the Materials and Process Simulation 33 of CH TFA to the transition states [TFACH ---H---Cl]·‡ and 3 2 Center (MSC) at Caltech. The computers used were funded from 34 ‡ [TFACH2---H---Br]· , the charge on TFA is increased from DURIP grants to the MSC. 35 −0.27 to −0.16 and −0.07, which is quite destabilizing due to 36 the electron-withdrawing nature of the TFA substituent. The ABBREVIATIONS ‡ 37 charge on Cl for CH3Cl and [ClCH2---H---Cl]· similarly increases from −0.12 to 0.10, but this is less destabilizing due BDE, bond dissociation energy; DFT, density functional theory; 38 OxE, oxy-esterification; OAc, acetate; PBF, Poisson-Boltzmann 39 to the greater polarizability of Cl. As for CH2Cl2 and [Cl2CH--- ‡ Finite element method; SCF, self-consistent field; SMR, steam- 40 H---Cl]· , the average charge on the Cl’s increases from −0.01 methane reformation; TFA, trifluoroacetate. to 0.16. Thus, we see the relative lack of over-oxidation 41 protection for CH Cl and CH Cl . 42 3 2 2 REFERENCES 43 This analysis explains the reason for the increased ΔG‡ of 1. General Chemistry: An Atoms First Approach; Halpern, J., Ed.; Howard University: Washington, 2014. 44 CH TFA activation, and thus the origin of protection against 3 2. Chapter 15.7: Fossil Fuels. In General Chemistry for 45 over-oxidation. Engineering; Halpern, J.; Sinex, S.; Johnson, S.; Eds.; Prince 46 George’s Community College: Largo, MD, 2016. 47 CONCLUSION 3. U.S. Energy Information Administration. Independent 48 Statistics & Analysis: Natural Gas. We have examined existing experimental data regarding the 49 https://www.eia.gov/dnav/ng/hist/rngwhhdm.htm (accessed oxidation of methane to methyl trifluoroacetate using a wide January 29, 2017). 50 variety of hypervalent iodine complexes, and used DFT to 4. Pachauri, R. K.; Reisinger, A. The Fourth Assessment Report 51 model viable mechanistic pathways for each reaction. Although of the Intergovernmental Panel on Climate Change; Working 52 there is a wide variety of iodine reagents capable of effecting Group 1: Geneva, Switzerland, 2007, Chapter 2. 5. The World Bank. Global Gas Flaring Reduction Partnership 53 the reaction, the tendency of iodine complexes to (GGFR). 54 disproportionate into I2 and I(V) serves as a “funnel” that unites http://www.worldbank.org/en/programs/gasflaringreduction 55 the sundry iodine precursors into the same mechanism based on (accessed January 29, 2017). 56 iodic acid. In our common iodate/OxE mechanism, iodic acid is 6. Cheng, W.-H.; Kung, H. H. Methanol Production and Use; − − 57 converted to IO2Cl2 or IOCl4 as the key intermediate, which Marcel Dekker: New York, NY, 1994; pp 283−317. 58 59 60 ACS Paragon Plus Environment The Journal of Physical Chemistry Page 10 of 12

7. ICIS. Methanol Prices, Markets, & Analysis. 34. Jang, W. J.; Shim, J. O.; Kim, H. M.; Yoo, S. Y.; Roh, H. S. 1 http://www.icis.com/chemicals/methanol/ (accessed January Catalysis Today 2019, 324, 15−26. 2 29, 2017). 35. Jocz, J. N.; Medford, A. J.; Sievers, C. Chemcatchem 2019, 8. National Renewable Energy Laboratory. Subcontract Report 11, 593−600. 3 NREL/SR-510-39943: Equipment Design and Cost 36. Liu, B. Y.; Oh, S. C.; Chen, H. Y.; Liu, D. X. J. Energy 4 Estimation for Small Modular Biomass Systems, Synthesis Chem. 2019, 30, 42−48. 5 Gas Cleanup, and Oxygen Separation Equipment. 37. Oda, A.; Ohkubo, T.; Yumura, T.; Kobayashi, H.; Kuroda, Y. 6 http://www.nrel.gov/docs/fy06osti/39943.pdf (accessed Inorg. Chem. 2019, 58, 327−338. 7 January 29, 2017). 38. Ouyang, M. Z.; Boldrin, P.; Maher, R. C.; Chen, X. L.; Liu, 9. Bal’zhinimaev, B. S.; Paukshtis, E. A.; Kovalev, E. V.; X. H.; Cohen, L. F.; Brandon, N. P. Appl. Catal. B 2019, 248, 8 Suknev, A. P.; Shalygin, A. S. (Uchrezhdenie RAN Institut 332−340. 9 Kataliza im. G. K. Boreskova SO RAN). Method of Selective 39. Rogozhnikov, V. N.; Snytnikov, P. V.; Salanov, A. N.; 10 Catalytic Oxychlorination of Methane to Methyl Chloride. Kulikov, A. V.; Ruban, N. V.; Potemkin, D. I.; Sobyanin, V. 11 RU Patent 2446881C2, April 10, 2012. A.; Kharton, V. V. Materials Letters 2019, 236, 316−319. 10. Aglulin, A. G. Kinet. Catal. 2009, 50, 427−434. 40. Sun, X.; Gao, Y. J.; Zhao, C. X.; Deng, S. W.; Zhong, X.; 12 11. Stauffer, J. E. Process for the Chlorination of Methane. Zhuang, G. L.; Wei, Z. Z.; Wang, J. G. Adv. Theo. Simul. 13 World Patent 9008117A1, July 26, 1990. 2019, 2, 2. 14 12. Rozanov, V. N.; Gvozd, E. V.; Kernerman, V. A.; Svetlanov, 41. Tian, Y. T.; Lu, Z.; Guo, X.; Wu, P. P. Intl. J. Electrochem. 15 E. B.; Trushechkina, M. A.; Treger, Y. A. Kinet. Catal. 1989, Sci. 2019, 14, 1093−1106. 30, 148−154. 42. Yan, X. L.; Hu, T.; Liu, P.; Li, S.; Zhao, B. R.; Zhang, Q.; 16 13. Riegel, H.; Schindler, H. D.; Sze, M. C. (CE Lummus). Jiao, W. Y.; Chen, S.; Wang, P. F.; Lu, J. J.; Fan, L. M.; 17 Oxychlorination of Methane. U.S. Patent 4,207,268A, June Deng, X. N.; Pan, Y. X. Appl. Catal. B 2019, 246, 221−231. 18 10, 1980. 43. Zhou, Y. Y.; Zhang, L.; Wang, W. Z. Nature Commun. 2019, 19 14. Labinger, J. A. J. Mol. Catal. 2004, 220, 27−35. 10. 20 15. Goldshlegger, N. F.; Tyabin, M. B.; Shilov, A. E.; Shteinman, 44. Hashiguchi, B. G.; Konnick, M. M.; Bischof, S. M.; A. A. Zh. Fiz. Khim. 1969, 43, 2174−2175. Gustafson, S. J.; Devarajan, D.; Gunsalus, N.; Ess, D. H.; 21 16. Goldshlegger, N. F.; Eskova, V. V.; Shilov, A. E.; Shteinman, Periana, R. A. Science 2014, 343, 1232−1237. 22 A. A. Zh. Fiz. Khim. 1972, 46, 1353−1354. 45. O’Reilly, M. E.; Fu, R.; Nielsen, R. J.; Sabat, M.; Goddard, 23 17. Shilov, A. E.; Shul'pin, G. B. Activation and Catalytic W. A. III; Gunnoe, T. B. J. Am. Chem. Soc. 2014, 136, 24 Reactions of Saturated Hydrocarbons in the Presence of 14690–14693. Metal Complexes; Kluwer Academic Publishers: Dordrecht, 46. Munz, D.; Webster-Gardiner, M. S.; Fu, R.; Strassner, T.; 25 2000; pp 259−317. Goddard, W. A. III; Gunnoe, T. B. ACS Catal. 2015, 5, 769– 26 18. Fekl, U.; Goldberg, K. I. Adv. Inorg. Chem. 2003, 54, 775. 27 259−320. 47. Webster-Gardiner, M. S.; Fu, R.; Fortman, G. C.; Nielsen, R. 28 19. Lersch, M.; Tilset, M. Chem. Rev. 2005, 105, 2471−2526. J.; Gunnoe, T. B.; Goddard, W. A. III. Catal. Sci. Technol. 29 20. Chepaikin, E. G. Russ. Chem. Rev. 2011, 80, 363−396. 2015, 5, 96–100. 21. Jones, C. J.; Taube, D.; Ziatdinov, V. R.; Periana, R. A.; 48. Fu, R.; O’Reilly, M. E.; Nielsen, R. J.; Goddard, W. A. III; 30 Nielsen, R. J.; Oxgaard, J.; Goddard, W. A., III. Angew. Gunnoe, T. B. Chem. Eur. J. 2015, 21, 1286–1293. 31 Chem., Int. Ed. 2004, 43, 4626−4629. 49. Webster-Gardiner, M. S.; Piszel, P. E.; Fu, R.; Chen, J.; 32 22. Periana, R. A.; Taube, D. J.; Gamble, S.; Taube, H.; Satoh, T.; McKeown, B. A.; Nielsen, R. J.; Goddard, W. A. III; Gunnoe, 33 Fujii, H. Science 1998, 280, 560−564. T. B. J. Mol. Catal. A: Chem. 2017, 426B, 381–388. 23. Periana, R. A.; Taube, D. J.; Taube, H.; Evitt, E. R. 50. Olah, G. A.; Gupta, B.; Felberg, J. D.; Ip, W. M.; Husain, A.; 34 (Catalytica, Inc.). Catalytic Process for Converting Lower Karpeles, R.; Lammertsma, K.; Melhotra, A. K.; Trivedi, N. J. 35 Alkanes to Esters, Alcohols, and to Hydrocarbons. U.S. J. Am. Chem. Soc. 1985, 107, 7097−7105. 36 Patent 5,306,855A, April 26, 1994. 51. Ding, K.; Metiu, H.; Stucky, G. D. ACS Catal. 2013, 3, 37 24. Kao, L. C.; Hutson, A. C.; Sen, A. J. Am. Chem. Soc. 1991, 474−477. 113, 700−701. 52. Dohi, T.; Kita, Y. ChemCatChem 2014, 6, 76−78; d. Moteki, 38 25. Ahlquist, M.; Nielsen, R. J.; Periana, R. A.; Goddard III, W. S. A.; Usui, A.; Zhang, T.; Solorio Alvarado, C. R.; Maruoka, 39 A. J. Am. Chem. Soc. 2009, 131, 17110-17115. K. Angew. Chem., Int. Ed. 2013, 52, 8657−8660. 40 26. Fu, R.; Nielsen, R. J.; Goddard, W. A. III; Fortman, G. C.; 53. Ding, K.; Zhang, A.; Stucky, G. D. ACS Catal. 2012, 2, 41 Gunnoe, T. B. ACS Catal. 2014, 4, 4455−4465. 1049−1056. 42 27. King, C. R.; Rollins, N.; Holdaway, A.; Konnick, M. M.; 54. Buddrus, J.; Plettenberg, H. Angew. Chem., Int. Ed. Engl. Periana, R. A.; Ess, D. H. Organometallics, 2018, 37, 3045– 1976, 15, 436. 43 3054. 55. Fortman, G. C.; Boaz, N. C.; Munz, D.; Konnick, M. M.; 44 28. Koppaka, A.; Park, S. H.; Hashiguchi, B. G.; Gunsalus, N. J.; Periana, R. A.; Groves, J. T.; Gunnoe, T. B. J. Am. Chem. 45 King, C. R.; Konnick, M. M.;Ess, D. H.; Periana, R. A. Angew. Soc. 2014, 136, 8393−8401. 46 Chem., Int. Ed. 2019, 58, 2241−2245. 56. Kalman. S. E.; Munz, D.; Fortman, G. C.; Boaz, N. C.; 29. Alvarez-Galvan, C.; Melian, M.; Ruiz-Matas, L.; Eslava, J. Groves, J. T.; Gunnoe, T. B. Dalton Trans., 2015, 44, 47 L.; Navarro, R. M.; Ahmadi, M.; Cuenya, B. R.; Fierro, J. L. 5294−5298. 48 G. Frontiers in Chemistry 2019, 7. 57. Schwartz, N. A.; Boaz, N. C.; Kalman, S. E.; Zhuang, T.; 49 30. Banke, K.; Hegner, R.; Schroder, D.; Schulz, C.; Atakan, B.; Goldberg, J. M.; Fu, R. Nielsen, R. J.; Goddard, W. A. III; 50 Kaiser, S. A. Fuel 2019, 243, 97−103. Groves, J. T.; Gunnoe, T. B. ACS Catal. 2018, 8, 3138–3149. 51 31. Brayko, A. S.; Shigarov, A. B.; Kirillov, V. A.; Kireenkov, V. 58. Bochevarov, A. D.; Harder, E.; Hughes, T. F.; Greenwood, J. V.; Kuzin, N. A.; Sobyanin, V. A.; Snytnikov, P. V.; Kharton, R.; Braden, D. A.; Philipp, D. M.; Rinaldo, D.; Halls, M. D.; 52 V. V. Materials Letters 2019, 236, 264−266. Zhang, J.; Friesner, R. A. Intl. J. Quant. Chem. 2013, 113, 53 32. Dedov, A. G.; Shlyakhtin, O. A.; Loktev, A. S.; Mazo, G. N.; 2110–2142. 54 Malyshev, S. A.; Moiseev, II. Doklady Chemistry 2019, 484, 59. Zhao, Y.; Truhlar, D. G. Theor. Chem. Account 2008, 120, 55 16−18. 215–241. 33. Fan, L. S.; Cheng, D. G.; Song, L. N.; Chen, F. Q.; Zhan, X. 60. Grimme, S.; Antony, J.; Erlich, S.; Krieg, H. J. Chem. Phys. 56 L. Chem. Eng. J. 2019, 369, 522−528. 2010, 132, 154104. 57 58 59 60 ACS Paragon Plus Environment Page 11 of 12 The Journal of Physical Chemistry

61. Hehre, W. J.; Ditchfield, R.; Pople, J. A. J. Chem. Phys. 1972, 78. Boswijk, K. H. ; Wiebenga, E. H. Acta Crystallogr. 1954, 7, 1 56, 2257–2261. 417–423. 2 62. Francl, M. M.; Pietro, W. J.; Hehre, W. J.; Binkley, J. S.; 79. Konnick, M. M.; Hashiguchi, B. G.; Devarajan, D.; Boaz, N. Gordon, M. S.; Defrees, D. J.; Pople, J. A. J. Chem. Phys. 1982, C.; Gunnoe, T. B.; Groves, J. T.; Gunsalus, N.; Ess, D. H.; 3 77, 3654–3665. Periana, R. A. Angew. Chem., Int. Ed. 2014, 53, 10490−10494. 4 63. Peterson, K. A. J. Chem. Phys. 2003, 119, 11099–11112. 80. Grant, D. J.; Garner, E. B. III; Matus, M. H.; Nguyen, M. T.; 5 64. Peterson, K. A.; Figgen, D.; Goll, E; Stoll, H.; Dolg, M. J. Peterson, K. A.; Francisco, J. S.; Dixon, D. A. J. Phys. Chem. 6 Chem. Phys. 2003, 119, 11113–11123. A, 2010, 114, 4254–4265. 7 65. Metz, B.; Stoll, H.; Dolg, M. J. Chem. Phys. 2000, 113, 2563– 81. Sememov, N. N. Some Problems in Chemical Kinetics and 2569. Reactivity; Princeton University Press: Princeton, NJ, 2017; p. 8 66. Tannor, D. J.; Marten, B.; Murphy, R.; Friesner, R. A.; Sitkoff, 191. 9 D.; Nicholls, A.; Ringnalda, M. N.; Goddard, W. A. III; Honig, 82. Carey, F. A.; Sundberg, R. J. Advanced Organic Chemistry, 10 B. J. Am. Chem. Soc. 1994, 116, 11875–11882. Part A: Structure and Mechanisms, 5th ed.; Springer: Berlin, 11 67. Marten, B.; Kim, K.; Cortis, C.; Friesner, R. A.; Murphy, R. 2008. B.; Ringnalda, M. N.; Sitkoff, D.; Honig, B. J. Phys. Chem. 83. Eyring, H. J. Chem. Phys. 1935, 3, 107–115. 12 1996, 100, 11775–11788. 84. Hammett, L. P. Physical Organic Chemistry, 1st ed.; McGraw- 13 68. Clark, T.; Chandrasekhar, J.; Spitznagel, G. W.; Schleyer, P. Hill: New York and London, 1940. 14 V. J. Comput. Chem. 1983, 4, 294–301. 85. Wiberg, N.; Holleman, A. F. Holleman-Wiberg’s Inorganic 15 69. Krishnan, R.; Binkley, J. S.; Seeger, R.; Pople, J. A. J. Chem. Chemistry; Academic Press: Cambridge, 2001. Phys. 1980, 72, 650–654. 86. Dasent, W. E.; Waddington, T. E. J. Chem. Soc. 1960, 16 70. Jaguar 8.4 User Manual, page 357, see note 86. 3350−3356. 17 71. Harris, F. E.; O’Konski, C. T. J. Am. Chem. Soc. 1954, 76, n+ 18 4317–4318. 87. For (IO)n specifically, we optimized the geometry in solution 19 72. Chase, M. W. Jr. NIST-JANAF Themochemical Tables, Fourth phase, since a gas phase optimization would inevitably fail due 20 Edition, J. Phys. Chem. Ref. Data, Monograph 9, 1998, 1– to the large amount of positive charge. 1951. 88. n = 2: ΔH = 19.9, ΔG = 30.0; n = 3: ΔH = 16.9, ΔG = 29.9; n = 21 73. Speight, J. Lange's Handbook of Chemistry, 17th ed.; McGraw- 4: ΔH = 18.1, ΔG = 32.8. 22 Hill: Boston, 2016; pp. 176–187. 89. n = 0: ΔH = −24.8, ΔG = −6.1; n = 1: ΔH = −23.9, ΔG = −4.2; 23 74. Michalkiewicz, B.; Jarosińska, M.; Łukasiewicz, I. Chem. Eng. n = 2: ΔH = −28.8, ΔG = −7.5; n = 3: ΔH = −27.9, ΔG = −5.8; 24 J. 2009, 154, 156−161. n = 5: ΔH = −28.7, ΔG = −8.3. 75. Gang, X.; Zhu, Y.; Birch, H.; Hjuler, H. A.; Bjerrum, N. J. 90. n = 0: ΔH = 0, ΔG = 0; n = 1: ΔH = 0.9, ΔG = 1.9; n = 2: ΔH = 25 Appl. Catal., A 2004, 261, 91−98. −4.1, ΔG = −1.4; n = 3: ΔH = −3.2, ΔG = 0.3; n = 5: ΔH = −3.9, 26 76. Periana, R. A.; Mirinov, O.; Taube, D. J.; Gamble, S. Chem. ΔG = −2.2. 27 Commun. 2002, 2376−2377. 91. Carey, F.A. Organic Chemistry, 6th Ed.; McGraw Hill: New 28 77. Bjerrum, N. J.; Xiao, G.; Hjuler, H. A. (Statoil Research York, 2006. 29 Centre). Process and Catalyst System for the Oxidation of 92. Viehe, H. G.; Janousek, Z; Merényi, R. Substituent Effects in Gaseous Hydrocarbons into Alcohols. World Patent Radical Chemistry; D. Reidel: Dordrecht, Holland, 1986. 30 9924383A1, May 20, 1999. 31 32 33 34 35 - 36 IO3 I Cl CH4 CH3TFA 37 (IO2)2S2O7 - Cl - Cl 38 IO4 I2 TFAH 39 I2O4 40 O (IO)2SO4 O 41 42 Methane oxidation by hypervalent iodine! 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 ACS Paragon Plus Environment The Journal of Physical Chemistry Page 12 of 12

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 12 60 ACS Paragon Plus Environment