Effect of Substituents on the Stability of Sulfur-Centered Radicals - The

Article pubs.acs.org/JPCA

Effect of Substituents on the Stability of Sulfur-Centered Radicals Published as part of The Journal of Physical Chemistry virtual special issue “Mark S. Gordon Festschrift”. Isa Degirmenci*,†,‡ and Michelle L. Coote*,‡ †

Chemical Engineering Department, Ondokuz Mayıs University, Samsun 55139, Turkey ARC Centre of Excellence for Electromaterials Science, Research School of Chemistry, The Australian National University, Canberra, ACT 2601, Australia

‡

S Supporting Information *

ABSTRACT: High-level ab initio calculations have been used to calculate the standard and inherent radical stabilities (RSEs) of a test set of 41 sulfur-centered radicals, chosen for their relevance in fields as diverse as combustion, atmospheric chemistry, polymer chemistry, and biochemistry. Radical stability was shown to be profoundly affected by substituents, varying over a 30 kcal mol−1 range for the test set studied. Like carbon-centered radicals, substituent effects on sulfur-centered radical stabilities result from the competition between the stabilizing effect of electron delocalization by lone pair donation and π-acceptance, and the destabilizing effect of σ withdrawal. However, in contrast to carbon-centered radicals, the heavier thiyl radicals are better able to undergo resonance and lone-pair donor interactions with heavier substituents. In particular, sulfur-containing lone pair donor and π-acceptor substituents have the greatest stabilizing effect, whereas σ-withdrawing substituents such as carbonyls and pyridines are the least stabilizing. The stabilities predicted using the standard definition and Zavitsas’s inherent RSEZ scheme are shown to be in surprisingly good agreement with one another for most species tested. The RSEZ values have also been shown to be capable of making chemically accurate estimates of bond energies by comparing our calculated values with 34 currently available experimental ones.

■

INTRODUCTION Sulfur-centered radicals play a crucial role in many important chemical reactions including chain transfer in radical polymerization, thiol−ene reactions, the thermal cracking of organic compounds, the vulcanization of rubber, combustion, and many biological processes.1 Despite their importance, their stability and reactivity has received only limited attention. Previous structure−reactivity studies have examined the barrier heights in hydrogen atom abstraction from thiols by carbon-centered radicals, a process that is both rapid and highly exothermic due to a combination of polar effects and the high relative stability of the product thiyl radical.2 Interestingly, despite their apparent stability, thiyl radical addition to alkenes is also rapid, though admittedly often reversible.3−6 This surprisingly high reactivity has also been attributed to polar factors, though a recent study suggests that such reactions also have low intrinsic barriers due to the superior ability of the sulfurcentered unpaired electrons to undergo resonance with the high-energy π* orbitals of the incoming substrate.7 Despite these important kinetic studies, the influence of substituents on the inherent stability of sulfur-centered radicals has not been extensively studied. An understanding of such structure− reactivity trends not only is useful in reagent design but also may aid a better understanding of oxidative stress and repair in biological systems. © XXXX American Chemical Society

Inherent radical stability (or radical stabilization energy) is a slippery concept, in the sense that it can only be defined in the context of a balanced chemical reaction, and yet to be useful it needs to be a transferable property of the radical itself that can be used to make qualitative predictions of structure reactivity trends in other types of chemical reactions.8 The most commonly used definition of radical stabilization energy for carbon-centered radicals (R•) is the enthalpy of the reaction 1.9,10 R• + H−CH3 → R−H + •CH3

(1)

In using this reaction to measure inherent radical stability, it is implicitly assumed that the inherent strengths of the C−H bonds in H−CH3 and R−H are sufficiently similar that trends in reaction energy reflect the relative ability of the substituents in R• to stabilize the radical. However, this equation does not take account of potential polar effects of the bond energy of the reference compounds, and this becomes a major problem noncarbon-centered radicals.11 Zipse has suggested that this may be mitigated using HO−H as a reference for RO−H bonds, HS−H for RS−H, and so forth.12 Though this makes Received: August 15, 2016 Revised: September 1, 2016

A

DOI: 10.1021/acs.jpca.6b08223 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A the reference value more sensible for each class of radicals, it still assumes that in these more polar RO−H or RS−H bonds, the substituents in R do not significantly influence the trends in the bond energies. A potential solution to this problem is offered by Zavitsas,13 who argued that one could explicitly correct for polar effects in bond energies using Pauling’s electronegativity equation, and in this way also allow for direct comparison of different types of radical,11 something not afforded by other approaches. In this definition of RSE (which we will call RSEZ to distinguish it from the traditional RSE) eq 2 is used: RSEz = 0.5 × (D[CH3 − CH3] − D[R − R])

Scheme 1. Structures in This Study

(2)

Using R−R bond energies (which we signify as D[R−R]), one rules out polar effects but potentially introduces steric effects in the R−R bond instead. This problem is then addressed by replacing the actual D[R−R] with a hypothetical “strain-free” D[R−R]calc which is obtained by applying Pauling’s electronegativity equation to the BDEs of CH3−R and Cl−R (both chosen to have minimal steric effects). D[CH3−R] =

1 (D[CH3−CH3] + D[R−R]calc ) 2 + 23(χ[CH3] − χ[R])2

D[Cl−R] =

(3)

1 (D[Cl−Cl] + D[R−R]calc ) 2 + 23(χ[Cl] − χ[R])2

(4)

In this equation, the χ values are Pauling’s electronegativity constants. Using known values of χ[•CH3] = 2.520, χ[Cl•] = 3.176, D[CH3−CH3] = 88.6 kcal mol−1, D[Cl−Cl] = 57.2 kcal mol−1, and calculated (or measured) values of D[CH3−R] and D[Cl−R] these two equations can be solved simultaneously to obtain the two unknowns: the electronegativity of R, χ[R], and its “strain-free” D[R−R]calc. This latter value can then be used in eq 2 to obtain RSEZ. Studies have shown that the RSEZ values (and associated χ values) obtained in this manner can be used to predict bond energies and enthalpies of formation for new compounds, as formed by new combinations of these component radicals.11,14 In the present work, we use this improved definition of inherent radical stability to study structure−reactivity trends in sulfur-centered radicals. We also compare the results with the trends given via the standard definition, eq 1. For simplicity we have not systematically shifted them by replacing CH4 with H2S as the reference, as this only affects the absolute numbers and not the relative trends. The test set selected for this study is provided in Scheme 1 and is designed to cover a broad range of potential substituent effects and include radicals of specific importance in vulcanization,15 combustion,16 atmospheric chemistry,17 initiation18 and chain transfer19,20 in free radical polymerization, self-healing reactions in dithiuram-based selfhealing polymers,21 and oxidative damage and repair in peptides.22

performed at this level to confirm the nature of stationary points and to calculate zero-point vibration energy and thermal corrections to the enthalpy (at 298.15 K) using the standard textbook formulas for an ideal gas under the harmonic oscillator/rigid rotor approximation. Improved energies were obtained using the G3(MP2)-RAD composite method, a highlevel composite ab initio method designed to approximate CCSD(T) calculations with a triple-ζ basis set via additivity approximations.26 To assess the accuracy of the computational methodology, the results were benchmarked against available experimental BDEs from literature (Figure 1 and Tables S1−2 of the Supporting Information). The mean absolute deviation (MAD) of calculated results from experiment is 2.3 kcal mol−1. From Figure 1, there are only two large outliers. These are for R = C6H5−C(S) and R = pNH2−C6H5, and their absolute deviations are 12.3 and 6.9 kcal mol−1, respectively. However, there is no direct experimental data for the former structure, as it was generated from a correlation. Among the other outliers, are the BDEs of HSS−SSH and C6H5−S−Cl; however, once these were recalculated with experimental heats of formation, the obtained result was in agreement with the theoretically calculated ones. Thus, with these exceptions, which are likely due to experimental error, the chosen methodology agrees with experiment to within quoted experimental error bars. Moreover, due to systematic error cancellation, the agreement between experimental and computational RSEs is even closer (MAD = 1.3−1.6 kcal mol−1, Table S3).

■

COMPUTATIONAL PROCEDURES All reaction enthalpies in this study were calculated using the Gaussian 0923 and Molpro 2012.24 program packages. All geometry optimizations and conformational analyses were performed at the M06-2X/6-31G(d) level of theory25 and complete conformational searches were performed at this level to identify global minima. Frequency calculations were also B

DOI: 10.1021/acs.jpca.6b08223 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

Figure 3. Dihedral angles of H−S−S−H, CH3−S−S−CH3, H−O− O−H, and CH3−O−O−CH3.

In addition to this, there are some larger outliers in the Figure 2, which require more explanation. One obvious noticeable outlier is CH3S(O)2−S(O)2CH3 that is destabilized by 10.9 kcal mol−1. This is in contrast to the other molecules studied, which are stabilized versus RSEcalc. As explained previously,11 the presence of electron-withdrawing oxygens on the sulfone sulfurs leads creates partial positive charges on the adjacent sulfur atoms, causing electrostatic repulsion. For dihydroxyethyl, diethyl, dithiocarbamate, D[RS−SR] exceeds D[RS−SR]calc by 14.1 kcal mol−1. This is because the OH groups have hydrogen bonding ability, which leads to further stabilization. In other examples, such as R = phenyl, pCl-phenyl, pCH3-phenyl, pCH3O-phenyl, and pNH2-phenyl, π-stacking interactions lead to stabilization to RS−SR. In the case of R = pNO2-phenyl, the delocalization of electrons from S atom to NO2 group through the phenyl ring enhances the interaction with lone pair electrons of one sulfur atom and σS−R * orbital of the other sulfur atom, leading to enhanced stabilization. Comparison of RSE and RSEZ. Figure 4 compares the standard RSE with RSEZ. It is worth noting that the RSEZ and

Figure 1. Comparison of experimental and theoretical (G3(MP2)RAD//M06-2X/6-31G(d)) bond dissociation energies (kcal mol−1) of RS−SR, RS−CH3, RS−Cl, and RS−H bonds. The line y = x is shown for clarity.

■

RESULTS AND DISCUSSION Bond dissociation energies for RS−X where X = H, Cl, CH3, and RS were calculated and are provided in Table S2 of the Supporting Information. These in turn were used to calculate the standard RSE and RSEZ, along with the strain-free D[RS− SR]calc and χ(RS•) values, as per eqs 1−4 (Table S2). All structures are provided in the Supporting Information. The various results are now presented and discussed in turn. Inherent Bonding Ability. The calculated inherent bonding ability, D[RS−SR]calc, systematically underestimates the D[RS−SR] for this set of molecules by an average of 7.5 kcal/mol (Figure 2). Because D[RS−SR]calc is free from any

Figure 2. Comparison of directly calculated bond dissociation energies (kcal/mol) D[RS−SR] versus predicted D[RS−SR]calc. The line y = x is shown for clarity, the MAD is 7.5 kcal/mol.

Figure 4. Comparison of the standard RSE with RSEZ. The line on the chart is y = x, but the line of best fit has the equation RSE = 1.1121RSEZ − 0.1062, and RSE and RSEZ have a correlation coefficient R of 0.94.

effect of inherent interactions, this deviation provides a measure of the interactions between the RS groups in the RSSR species. It is well-known that hydrogen disulfide (H−S−S−H) prefers a nonlinear structure like hydrogen peroxide (H−O−O−H) due to repulsion between lone pair electrons of adjacent sulfur atoms (Figure 3). In this preferred configuration the lone pair orbitals are poised to interact in a stabilizing manner with the opposite σ*S−H (or more generally σ*S−R in R−S−S−R). As a result of this interaction, the actual D[RS−SR] values are higher than the D[RS−SR]calc values, which lack this stabilizing contribution.

RSE values are in surprisingly good absolute agreement with one another. The correlation coefficient is 0.94, and the two main outliers are the sulfonyl CH3S•(O)2 and sulfinyl CH3S•(O) radicals which, as noted earlier, would introduce destabilization into the reference compound (in the standard RSE) not present in the RSEZ values, which are designed to avoid this. The remaining species are in generally very good agreement between the two schemes, indicating that the effects of R on the RS−H bond strength in the standard RSE are relatively minor compared with the corresponding effects on the radical. C

DOI: 10.1021/acs.jpca.6b08223 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

heteroatom-centered radical (cf. HS• is itself higher than CH3• by 13.4 kcal mol−1). In a similar vein conjugated carbonyl groups stabilize carbon-centered radicals via π-acceptance, but for thiyl radicals these effects are diminished and their σwithdrawing properties are more dominant. Thus, for sulfurcentered radicals, the most effective π-acceptor groups are the thiocarbonyl substituents. Likewise, the most effective lone pair donor substituents are sulfurs (rather than O or N), as these are on the same row of the periodic table. Moreover, the correlation between spin density and radical stability is not perfect and clearly other effects exist. In particular, is well-known that carbon-centered radicals are destabilized by adjacent electronegative or electron-withdrawing substituents such as, CF3, CCl3, NO2, F, and OC(O)R whereas they are stabilized by electron-donating substituents such as NH2.10,12 In the sulfur-centered radicals we see similar trends: for instance, replacing O with the less electronegative S in the CH3C(O)S• radical leads to 13 kcal mol−1 of additional stabilization. In the ester analogues, replacing both O atoms with S, increases the RSE by approximately 10 kcal mol−1. In a related vein, the RSEZ values increase from 23.0 to 26.2 kcal mol−1 for HC(S)S• to CH3C(S)S• radicals, and this can be explained by the electron-donating ability of the R = CH3 group over H with the conjugated π orbital system in the −C(S)SR. This effect is further enhanced with the electron-donating nitrogen substituents in (CH3CH2) (HOCH2CH2)NC(S)S•, (CH3CH2)2NC(S)S•, and (CH3)2NC(S)S• The effect of electronegativity on the stability of the radicals is also seen in the effect of para-substituents in the p-X−C6H4−S• series. RSEZ values increase as X = NO2 < CH3O < H < Cl < CH3 < NH2, which reflects the electron-donating ability of the substituents. The radicals with heterocyclic moieties attached to the sulfur-centered radical have also been considered in this study due to their importance as iniferter or photoinitiators for free-radical polymerization. Among the heterocyclic systems the imidazole substituents are relatively stabilizing, and the pyridine substituents are not, whereas the triazole substituents display intermediate properties. This reflects a competition between electron delocalization on the one hand and σ withdrawal on the other. Bond Energy Prediction Using Radical Stabilities. If radical stabilization energy values are an inherent and transferable property of the radical, they should be able to predict (at least qualitatively) new reaction energies that were not used in their original determination. For instance, it was recently shown that the RSEZ and χ values for a large test set of carbon and heteroatom-centered radicals could be used to predict the bond dissociation energies of new compounds. That is, using the values of RSEZ and χ for R• and R′•, one could combine them in new ways to predict values of D[R−R′] via eq 5.11

Effects of Substituents on Radical Stability. Figure 5 shows a bar chart of the RSE and RSEZ values, organized from

Figure 5. Stabilities of sulfur-centered radicals (Table S2, Supporting Information).

least to most stable based on RSEZ; the corresponding raw data are provided in Table S2 of the Supporting Information. The substituents R have an enormous effect on radical stability, covering a range of nearly 30 kcal mol−1 for the substituents studied. As noted above, with the exception of the sulfonyl and sulfinyl radicals, the trends in RSE and RSEZ are very similar. Nonetheless, remainder of this discussion is based on the RSEZ trends, which are expected to better reflect inherent radical stability.11 As in the case of carbon-centered radicals,10,11 the primary stabilizing effect of R is to delocalize the unpaired electron and as such the correlation between radical stability and spin density on the sulfur is relatively high (R = 0.9, Figure S1 of the Supporting Information). As a result, the sulfur-centered radicals are stabilized by conjugated π-accepting or lone-pair donor groups. However, it is important to note that some seemingly conjugated π-acceptor groups do not significantly delocalize or stabilize the sulfur-centered radicals. For instance, CH2CHS• has one of the lowest RSEZ values among those studied (13.8 kcal mol−1) and a relatively localized spin density (0.993). This is because delocalizing the unpaired electron onto the π-system involves participation of a resonance contributor with the unpaired electron on carbon rather than sulfur (i.e., CH2•CHS), which is inherently higher in energy than the

D[R−R′]calc = D[CH3−CH3] − RSE Z[R•] − RSE Z[R′• ] + 23(χ[R] − χ[R′])2

(5)

In doing this, it needs to be recognized that the D[R−R′]calc values are the hypothetical “strain free” values in which the interactions between R and R′ are the polar effects accounted for by Pauling’s equation. As discussed above, in compounds where there is significant steric crowding, resonance interactions across the R−R′ bond, or other specific interactions such hydrogen bonding, the values will differ and this difference can be used to study those interactions. D

DOI: 10.1021/acs.jpca.6b08223 J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A

■

Notwithstanding those limitations, the previous study, the MAD of all predicted data was less than 2 kcal mol−1, which is was within the experimental error range. The same approach is used here to predict 34 D[R−R′] energies formed both from mixed combinations of thiyl radicals from the present work and from combinations of these thiyl radicals with carbon-centered radicals from ref 11. The results are tabulated in the Table S4 of the Supporting Information and a comparison of the predicted bond energies and experiment is provided in Figure 6. We obtained an MAD of

Article

ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpca.6b08223. Further computational data including all computed and experimental bond energies, electronegativities and RSEs, correlation between spin density and radical stability, complete optimized geometries and associated total energies, entropies, enthalpies, and Gibbs free energies (PDF)

■

AUTHOR INFORMATION

Corresponding Authors

*I. Degirmenci. E-mail: [email protected]. Phone: (+90) 3623121919, extension 1533. *M. L. Coote. E-mail: [email protected]. Phone: (+61) 2 6125 3771. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS I.D. gratefully acknowledges the Scientific and Technological Research Council of Turkey (TUBITAK) under 2219 grant and Naomi Haworth for her valuable support. M.L.C gratefully acknowledges generous allocations of supercomputing time on the National Facility of the Australian National Computational Infrastructure and financial support from the Australian Research Council Centre of Excellence for Electromaterials Science.

Figure 6. Comparison of the experimental D[R−R′] and predicted D[R−R′]calc values (kcal mol−1). The line on the chart is y = x, and the mean absolute deviation is 2.4 kcal mol−1.

■

REFERENCES

(1) Deńeś, F.; Pichowicz, M.; Povie, G.; Renaud, P. Thiyl Radicals in Organic Synthesis. Chem. Rev. 2014, 114, 2587−2693. (2) Beare, K. D.; Coote, M. L. What Influences Barrier Heights in Hydrogen Abstraction from Thiols by Carbon-Centered Radicals? A Curve-Crossing Study. J. Phys. Chem. A 2004, 108, 7211−7221. (3) Walling, C.; Helmreich, W. Reactivity and Reversibility in the Reaction of Thiyl Radicals with Olefins. J. Am. Chem. Soc. 1959, 81, 1144−1148. (4) Lalevée, J.; Allonas, X.; Fouassier, J. P. Search for High Reactivity and Low Selectivity of Radicals toward Double Bonds: The Case of a Tetrazole-Derived Thiyl Radical. J. Org. Chem. 2006, 71, 9723−9727. (5) Lalevée, J.; Allonas, X.; Morlet-Savary, F.; Fouassier, J. P. Respective Contributions of Polar vs Enthalpy Effects in the Addition/ Fragmentation of Mercaptobenzoxazole-Derived Thiyl Radicals and Analogues to Double Bonds. J. Phys. Chem. A 2006, 110, 11605− 11612. (6) Northrop, B. H.; Coffey, R. N. Thiol−Ene Click Chemistry: Computational and Kinetic Analysis of the Influence of Alkene Functionality. J. Am. Chem. Soc. 2012, 134, 13804−13817. (7) Degirmenci, I.; Coote, M. L. Comparison of Thiyl, Alkoxyl, and Alkyl Radical Addition to Double Bonds: The Unusual Contrasting Behavior of Sulfur and Oxygen Radical Chemistry. J. Phys. Chem. A 2016, 120, 1750−1755. (8) Coote, M. L.; Dickerson, A. B. The Measurement and Meaning of Intrinsic Radical Stability: Are Chemical Questions just problems in Applied Mathematics? Aust. J. Chem. 2008, 61, 163−167. (9) Griller, D.; Ingold, K. U. Persistent Carbon-centered Radicals. Acc. Chem. Res. 1976, 9, 13−19. (10) Coote, M. L.; Lin, C. Y.; Beckwith, A. L. J.; Zavitsas, A. A. A Comparison of Methods for Measuring Relative Radical Stabilities of Carbon-centred Radicals. Phys. Chem. Chem. Phys. 2010, 12, 9597− 9610. (11) Coote, M. L.; Lin, C. Y.; Zavitsas, A. A. Inherent and Transferable Stabilization Energies of Carbon- and Heteroatom-

2.4 kcal mol−1 for this set of structures. As noted above, this prediction assumes there is no significant interaction between the two fragments in the R−R′ compounds, which is not always the case. Indeed, the four largest outliers all involve conjugative or hyperconjugative interactions between R and R′: HS−SCH3 (9.4 kcal mol−1); C2H5SC(O)CH3 (8.6 kcal mol−1); CH3S− CN (7.4 kcal mol−1); PhS−SSH (5.2 kcal mol−1). When these four are removed, the MAD drops to 1.7 kcal mol−1, and the average deviation drops even lower when other compounds suspected of steric or resonance interactions are removed. In other words, the RSEZ values are inherent properties of the radical and useful in making independent predictions.

■

CONCLUSIONS Substituents can affect the stability of sulfur-centered radicals by over 30 kcal mol−1. These effects can be effectively measured using either standard RSEs or Zavitsas’s universal RSE definition, with the latter shown to be capable of accurate predictions of new bond energies. Like carbon-centered radicals, substituent effects on sulfur-centered radical stabilities generally represent a competition between electron delocalization on the one hand and σ withdrawal on the other. However, due to its position on the periodic table, thiyl radicals are better able to undergo resonance and lone-pair interactions with heavier substituents than their carbon-centered radical counterparts. In particular, lone pair donation and π-acceptance involving sulfur-containing substituents have the greatest stabilizing effect, whereas σ-withdrawing substituents such as carbonyls and pyridines are the least stabilizing. E

DOI: 10.1021/acs.jpca.6b08223 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A centred Radicals on the Same Relative Scale and Their Applications. Phys. Chem. Chem. Phys. 2014, 16, 8686−8696. (12) Hioe, J.; Zipse, H. Radical Stability and its Role in Synthesis and Catalysis. Org. Biomol. Chem. 2010, 8, 3609−3617. (13) Matsunaga, N.; Rogers, D. W.; Zavitsas, A. A. Pauling’s Electronegativity Equation and a New Corollary Accurately Predict Bond Dssociation Enthalpies and Enhance Current Understanding of the Nature of the Chemical Bond. J. Org. Chem. 2003, 68, 3158−3172. (14) Coote, M. L.; Zavitsas, A. A. Using Inherent Radical Stabilization Energies to Predict Un-known Enthalpies of Formation and Associated Bond Dissociation Energies of Complex Molecules. Tetrahedron 2016, DOI: 10.1016/j.tet.2016.03.015. (15) See for example: Akiba, M.; Hashim, A. Vulcanization and Crosslinking in Elastomers. Prog. Polym. Sci. 1997, 22, 475−521. (16) See for example: Daum, K. A.; Massoglia, M. F.; Shendrikar, A. D. An Analysis of Sulfur Control Strategies for the Oil Shale Industry. J. Air Pollut. Control Assoc. 1982, 32, 391−392. (17) See for example: Hatakeyama, S.; Akimoto, H. Reactions of Hydroxyl Radicals with Methanethiol, Dimethyl Sulfide, and Dimethyl Disulfide in Air. J. Phys. Chem. 1983, 87, 2387−2395. (18) See for example: Ferington, T.; Tobolsky, A. Organic Disulfides as Initiators of Polymerization: Tetramethylthiuram Disulfide. J. Am. Chem. Soc. 1955, 77, 4510−4512. (19) See for example: Šebenik, A. Living Free-radical Block Copolymerization using thio-iniferters. Prog. Polym. Sci. 1998, 23, 875−917. (20) See for example: Valdebenito, A.; Encinas, M. Thiophenols as Chain Transfer Agents in the Polymerization of Vinyl Monomers. Polymer 2005, 46, 10658−10662. (21) See for example: Amamoto, Y.; Otsuka, H.; Takahara, A.; Matyjaszewski, K. Self-healing of Covalently Cross-linked Polymers byReshuffling Thiuram Disulfide Moieties in air under Visible Light. Adv. Mater. 2012, 24, 3975−3980. (22) See for example: Rauk, A.; Armstrong, D. A.; Fairlie, D. P. Is Oxidative Damage by β-amyloid and Prion Peptides Mediated by Hydrogen Atom Transfer from Glycine α-carbon to Methionine Sulfur within β-sheets? J. Am. Chem. Soc. 2000, 122, 9761−9767. (23) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; et al. Gaussian 09, revision D.01; Gaussian, Inc.: Wallingford, CT, 2009. (24) Werner, H.-J.; Knowles, P. J.; Knizia, G.; Manby, F. R.; Schütz, M.; Celani, P.; Korona, T.; Lindh, R.; Mitrushenkov, A.; Rauhut, G.; et al., MOLPRO 2012.1; 2015, http://www.molpro.net. (25) Zhao, Y.; Truhlar, D. G. The M06 Suite of Density Functionals for Main Group Thermochemistry, Kinetics, Noncovalent Interactions, Excited States, and Transition Elements: Two New Functionals and Systematic Testing of Four M06 Functionals and Twelve Other Functionals. Theor. Chem. Acc. 2008, 120, 215−241. (26) Henry, D. J.; Sullivan, M. B.; Radom, L. G3-RAD and G3XRAD: Modified Gaussian-3 (G3) and Gaussian-3X (G3X) procedures for radical thermochemistry. J. Chem. Phys. 2003, 118, 4849−4860.

F

DOI: 10.1021/acs.jpca.6b08223 J. Phys. Chem. A XXXX, XXX, XXX−XXX