Monitoring electrophilic intermediates in reactions of thiols in aqueous solution directly with ¹⁹F NMR

Abstract

Mechanistic studies of thiol reactivity can be challenging because electrophilic reaction intermediates, such as sulfenic acids (RSOH) and sulfenyl chlorides (RSCl), are generally too reactive to be observed directly. Herein we report the design and synthesis of a sterically-encumbered fluorinated triptycene thiol which enables direct observation of reaction intermediates in aqueous buffer by ¹⁹F NMR, as demonstrated in reactions with hydrogen peroxide and hypochlorous acid. Reactions with H₂O₂ resulted in the formation of a persistent RSOH species, which was subsequently converted to a sulfinic acid (RSO₂H) and then a sulfonic acid (RSO₃H), while RSCl was found to be the intermediate in reactions with HOCl. Utilizing the same scaffold, reactions of thiol with thermally and photochemically generated singlet oxygen afforded RSO₂H as the primary product. The stark difference in product profile from sterically-unencumbered thiols – which yield disulfides – implies that the reaction proceeds through a sulfenyl hydroperoxide (RSOOH) intermediate. Sulfenic acids, which were not observed in reactions of thiols with singlet oxygen, were also found to rapidly react with singlet oxygen to afford sulfinic acids, which is proposed to involve initial formation of an analogous sulfinyl hydroperoxide (RS(O)OOH). The formation and reactions of RSOOH are explored by computations. Use of the water-soluble fluorinated triptycene scaffold to probe reductive processes on RSOH (e.g., with ascorbate and/or iron) is also illustrated, wherein it was found that RSOH are surprisingly resistant to reductive heterolysis – in stark contrast with hydroperoxides – owing to their strong S–O bond.

---

1 Introduction

Thiol redox reactivity is central to many aspects of biology. Cysteine, the unique thiol-containing amino acid, provides proteins and peptides (i.e. glutathione) with unique catalytic,^1,2 structural,³ signaling,^4,5 and antioxidant⁶ functions. In many cases, these functions are underpinned by thiol/disulfide exchange reactions, while in others, oxidation of the thiol to a reactive intermediate is believed to be key. Of particular interest are the reactions of cysteine thiols with H₂O₂, the pervasive ‘reactive oxygen species’ that abounds the cell – and at elevated levels in pathological contexts. Disulfides are primarily formed in these reactions, but sulfinic acids (RSO₂H) and sulfonic acids (RSO₃H) can also be observed depending on conditions.⁷ All of these products are believed to result from the initial formation of a sulfenic acid (RSOH) from substitution of the thiolate on H₂O₂; the sulfenic acid then either reacts as an electrophile, undergoing substitution by another equivalent of thiolate to yield disulfide, or as a nucleophile, where it is oxidized by another equivalent of H₂O₂ to yield RSO₂H and then RSO₃H. While thiol oxidation to RSO₂H is reversible in the cell, RSO₃H formation is irreversible, abolishing the function of the thiol.⁸

The reactivity of RSOH as both electrophile and nucleophile leads to their rapid self-reaction, making isolation and determination of accurate physicochemical properties and/or reaction kinetics challenging – especially under physiologically-relevant conditions (i.e. in aqueous solutions).⁹ Although some protein sulfenic acids are stable and can be observed by mass spectrometry,¹⁰ the protein structure can make mechanistic studies and generalizations about RSOH reactivity difficult.^11,12 To provide insight to their reactivity, several persistent small-molecule RSOH have been prepared,^13,14 including a sterically-encumbered cysteine sulfenic acid.¹⁵ Unfortunately, the hydrophobicity of the scaffold restricted the study of its chemical properties to largely organic solutions (i.e. 20 : 1 THF/H₂O), which cannot be buffered for studies of the pH-dependence of the reactivity of thiols and sulfenic acids – a key consideration given that physiologically-relevant thiol chemistry is dominated by the more reactive conjugate base (thiolate), which is present to a substantial extent at pH 7.4 (thiol pK_as range from 5.0 to 8.8).¹⁶

We previously used a triptycene scaffold functionalized with a bridgehead fluorine atom to monitor the reaction of thiol and H₂O₂ in buffered methanol directly using ¹⁹F NMR.¹³ The triptycene moiety precludes electrophilic reactions of the sulfenic acid because a nucleophile cannot access the orbital localized on the sulfur atom, unequivocally confirming it as the initial product. This approach not only enabled detailed kinetic measurements and unambiguous characterization of the specific-base catalysis mechanism responsible, but also the successive oxidations to sulfinic and sulfonic acids which follow.¹⁷ Unfortunately, studies of the reaction of thiols and sulfenic acids with stronger oxidants, including other biologically-relevant ‘reactive oxygen species’ such as hypochlorous acid (HOCl) and singlet oxygen (¹O₂), were not possible due to their competitive reactions with the (amine-based) buffers. Similarly to the reactions of thiols with H₂O₂, reactions of HOCl and ¹O₂ have been shown to yield disulfides, but the mechanisms remain unclear due, in part, to the inability to directly observe reaction intermediates. We surmised that substitution of the 9-triptycenethiol scaffold to increase water solubility would enable direct observation of any intermediates formed in the reactions of thiols with HOCl, ¹O₂ and/or other high energy oxidants. Moreover, it would permit analogous studies on sulfenic acids and the investigation of other reactions which have been proposed to involve RSOH intermediates. Herein we describe the preparation of such a scaffold, its use to provide insights to the reactions of thiols and sulfenic acids with H₂O₂, HOCl and ¹O₂, and describe other applications to study sulfenic acid reactivity.

2 Results & discussion

2.1. Synthesis of thiol 1 and its oxidation products (6, 7, 8)

To improve the water-solubility of the fluorinated 9-triptycenethiol used in our previous work, we incorporated pendant tetraethylene glycol (TEG) units on the triptycene backbone as in 1. The synthesis of 1 was carried out in 7 steps from 9-bromoanthracene with an overall yield of 9% (Fig. 1A). Lithium-bromide exchange on 9-bromo-10-fluorotriptycene (2)¹⁷ and substitution of the resulting triptycenyl lithium on methyl disulfide afforded methyl sulfide 3 in 92% yield. Selective bromination of 3 was challenging; the Lewis base-catalyzed approach developed by Nishii and Miura¹⁸ was most successful, leading to a mixture of dibrominated products 4a (meta, meta to the bridgehead C–S bond) and 4b (meta, para). Interestingly, despite its striking similarity, the fluorinated triptycene 3 did not assist the reaction on its own and the TpSMe used by Nishii and Miura¹⁸ had to be added for the reaction to proceed to completion. Indeed, the presence of the fluorine atom also precluded access to the tribrominated triptycene, which was readily obtained under these reaction conditions in its absence. Both isomers of 4 were then subjected to Ullmann-type coupling conditions in the presence of Cu(OAc)₂ using TEG as both nucleophile and solvent, sluggishly affording a corresponding isomeric mixture of 5 in 78% yield. The methyl sulfide was then cleaved using freshly prepared lithium thioethoxide in absolute DMPU at 120 °C to obtain the target thiol 1 in 51% yield. Reverse phase HPLC could resolve 1a from 1b, enabling its unequivocal characterization, but most of our experiments were carried out on the mixture as it afforded a second data set in each independent experiment (vide infra).

Fig. 1 (A) Preparation of thiols 1, and (B) corresponding sulfenic acids 6, sulfinic acids 7 and sulfonic acids 8.

The sulfenic acids (6) and sulfinic acids (7) derived from 1 were prepared in 46% and 22% yield, respectively, from the direct oxidation of thiols 1 by H₂O₂ (2 eq.) in H₂O/THF (1 : 1) containing K₂HPO₄ and separated by flash column chromatography. The sulfonic acids (8) were prepared similarly by the overnight reaction of 1 with 5 eq. of H₂O₂ (Fig. 1B).

2.2. Reactions of 1 with H₂O₂

To obtain reaction kinetics for the oxidation of thiols 1 by H₂O₂ in water, 1 was treated with an excess of H₂O₂ in phosphate-buffered D₂O/THF (3 : 1, v/v) at pD 7.8 (pH 7.4),¹⁹ and their consumption was monitored by ¹⁹F NMR along with the corresponding appearance of product sulfenic, sulfinic and sulfonic acids (quantified relative to added 2,2,2-trifluoroethanol) as shown in Fig. 2. Since the two regioisomers of 1 and their corresponding sulfur oxyacids have distinct ¹⁹F NMR shifts, two data sets were collected per experiment (Fig. 2B). These data sets were fit to a simple kinetic model for successive bimolecular reactions of 1, 6 and 7 with H₂O₂ (Fig. 2C), from which observed rate constants (k_obs) of 0.10, 0.013 and 0.0014 M⁻¹ s⁻¹ were derived for the reactions of thiol, sulfenic and sulfinic acid, respectively. Not only is this the first example of these reactions being followed directly in aqueous solution but, as far as we are aware, these are the first rate constants for RSOH and sulfinic acid oxidation by H₂O₂ measured in aqueous solution.‡

Fig. 2 (A) Reaction scheme for oxidation of thiols 1 by H₂O₂, including pK_a values determined for each of 1, 6 and 7 and absolute rate constants obtained for the reactions of the conjugate bases of each of 1, 6, and 7 with H₂O₂ in 50 mM phosphate-buffered pD 7.8 D₂O/THF (3 : 1). The rate constants are derived from the mean of the individual rate constants determined for each of the a and b isomers (see ESI†). (B) Representative time-resolved ¹⁹F NMR spectra obtained upon treatment of 1 (5 mM) with 25 mM H₂O₂ in 50 mM phosphate-buffered pD 7.8 D₂O/THF (3 : 1). (C) Representative reaction profiles of 1a (2.8 mM) and 1b (2.2 mM) with 5 eq. H₂O₂ in pD 7.8 D₂O/THF (3 : 1), fits of the data to a simple kinetic model for successive bimolecular reactions of 1, 6 and 7 with H₂O₂ and the corresponding observed rate constants (in M⁻¹ s⁻¹) derived from the fits.

Previously measured k_obs for the reactions of the water-insoluble triptycene thiol, sulfenic acid and sulfinic acid with H₂O₂ in methanol at (ca. pH = 7.66 in H₂O)²⁰ were 2.5 × 10⁻³, 9.5 × 10⁻⁴ and 7.3 × 10⁻⁴ M⁻¹ s⁻¹, respectively. Given that these reactions are specific-base catalyzed (Fig. 2A), we suspected that the origin of the large solvent effects on the reactions of the thiol (40×) and sulfenic acid (13×) is the more favourable pre-equilibrium in aqueous solution relative to methanol. To derive absolute rate constants, the pK_as of thiol 1 (8.3 ± 0.1) and sulfenic acid 6 (10.4 ± 0.1) were measured in D₂O/THF (3 : 1) (the values for 1a and 6a were within the margin of error on the measurements made with the mixtures and corresponding values in H₂O afforded 8.0 ± 0.1 and 10.1 ± 0.1, respectively). Expectedly, deprotonation of thiol and sulfenic acid were found to be much more favourable in water than in methanol (pK_a = 11.6 and 12.8, respectively).¹⁷ With these pK_a values at hand, we derived absolute (pH-independent) rate constants for the reactions of thiolate (k = 0.43 M⁻¹ s⁻¹, compared to 0.13 M⁻¹ s⁻¹ in methanol) and sulfenate (k = 5.2 M⁻¹ s⁻¹, compared to 0.75 M⁻¹ s⁻¹ in methanol) with H₂O₂. As expected from the alpha effect²¹ and our previous data,¹⁷ sulfenate anions were found to be more reactive nucleophiles than thiolate anions. Sulfinate was found to be much less reactive as a nucleophile (k = 1.4 × 10⁻³ M⁻¹ s⁻¹ compared to 5.8 × 10⁻⁴ M⁻¹ s⁻¹ in methanol) consistent with the greater relative stability of the anion (pK_a ∼2¹⁷vs. 10 for the corresponding RSOH). Overall, the rate constant determined for 1 is toward the low end of the range of those which have previously been reported (derived from consumption of thiol(ate)s),§ which span from 0.16 M⁻¹ s⁻¹ (for N-acetylcysteine) to 4.5 M⁻¹ s⁻¹ (for penicillamine) at pH 7.4,¹² which is presumably due to the electron-withdrawing effect of the 10-fluorotriptycenyl scaffold.

2.3. Reactions of 1 with HOCl

At physiological pH, thiols react rapidly with HOCl (k > 10⁸ M⁻¹ s⁻¹),²² with sulfenyl chlorides being proposed as the initial products, but which are not observed due to their rapid reaction with thiols to yield disulfides.²³ We observed instant formation of a single product (13a) upon addition of limiting NaOCl to a solution of 1a in pD 7.8 D₂O/THF (Fig. 3A). The ¹⁹F NMR chemical shift (−205.9 ppm) did not correspond to the sulfenic acid 6a (−206.6 ppm), implying formation of the sulfenyl chloride. However, when thiol 1a was treated with N-chlorosuccinimide in chloroform, a product (14a) with a distinct ¹⁹F NMR chemical shift (−206.4 ppm) was obtained (Fig. 3B).²⁴ High-resolution mass spectrometry indicated that 14a was the sulfenyl chloride. While 14a was stable over extended periods in pD 7.8 D₂O/THF buffer, addition of 1a readily yielded 13a, which we have characterized as the disulfide of 1a (Fig. 3C). Since substitution of thiolate 1a on sulfenyl chloride 14a to form the disulfide is unlikely (the nucleophile cannot access the orbital localized on the sulfur atom), the thiolate presumably reacts with the sulfenyl chloride by electron transfer, resulting in the formation of thiyl radicals which dimerize to give the disulfide (Fig. 3B). An analogous mechanism has been proposed for the observation of disulfides from the reaction of sulfenyl chlorides and amines.²⁵ The intervention of thiyl radicals in the reaction of non-hindered thiols with HOCl has previously been demonstrated by Davies,²⁶ who characterized DMPO adducts of thiyl radicals by EPR, suggesting that this pathway is not unique to the reaction of 9-triptycenethiols.

Fig. 3 (A) Representative ¹⁹F NMR spectra obtained before and after treatment of 1a (17 mM) with 0.35 eq. of NaOCl in 50 mM phosphate-buffered pD 7.8 D₂O/THF (3 : 1). (B) Preparation of sulfenyl chloride 14a from 1a and representative ¹⁹F NMR in CDCl₃. (C) Representative ¹⁹F NMR spectra obtained before and after treatment of 14a (17 mM) with 0.65 eq. of 1a in 50 mM phosphate-buffered pD 7.8 D₂O/THF (3 : 1). (D) Proposed mechanism of disulfide formation upon thiol oxidation by hypochlorous acid.

2.4. Reactions of 1 with ¹O₂

Similarly to their reactions with H₂O₂ and HOCl, thiols have been shown to yield disulfides upon reaction with ¹O₂.²⁷ To provide insight to the mechanism of this reaction, we investigated the reaction of 1 with ¹O₂ derived from decomposition of the endoperoxide of sodium naphthalene-1,4-dipropionate (NDPO₂) at 37 °C (t_1/2 = 23 min).^28,29 The reactions were again monitored by ¹⁹F NMR at 37 °C in pD 7.8 phosphate buffered D₂O/THF (3 : 1, v/v), but now in the presence of 4 eq. of NDPO₂in lieu of either H₂O₂ or HOCl. After 33 minutes, the conversion of thiol was ∼45%, giving primarily sulfinates 7 (∼25%). Importantly, disulfide 13a was not observed. Other products included two presumed pairs of unknown isomers (∼10%) downfield of 8 at δ = −204.55 ppm and −205.25 ppm (Fig. 4A). Control experiments suggest that the formation of these unknowns can be attributed to a reaction occurring in THF in the presence of ¹O₂ (see ESI Fig. S6†). The missing ∼10% in the mass balance suggests that some signal suppression from the formation of persistent paramagnetic species is taking place (see ESI Fig. S3†).^30,31

Fig. 4 (A) Representative time-resolved ¹⁹F NMR spectra obtained upon treatment of 1 (5 mM) with 4 eq. of NDPO₂ in 50 mM phosphate-buffered pD 7.8 D₂O/THF (3 : 1). (B) Representative time-resolved ¹⁹F NMR spectra obtained of the reaction of 9 (20 mM) with photochemically generated ¹O₂ in CD₃OD/H₂O (15 : 1) containing 60 mM NaOH. (C) Scheme accounting for the different product distributions arising from reaction of hindered and unhindered thiols with ¹O₂. (D) Mechanistic possibilities and CBS-QB3 calculated free energies (298 K) of the reaction of a model (t-butyl) thiol with ¹O₂ (SET = single electron transfer; HAT = hydrogen atom transfer). Data calculated using a polarizable continuum model using the integral equation formalism (IEFPCM) parameterized for water are given; associated gas phase values are given in the ESI.†

Consistent with the foregoing experiments with 1, when thiol 9 was subjected to Rose Bengal-photosensitized oxygenation in 15 : 1 CD₃OD : H₂O containing 3 eq. NaOH, the corresponding sulfinate 11 and sulfonate 12 were formed (Fig. 4B) (see Fig. S4 of the ESI† for corresponding ¹H NMR spectra). Traces of disulfide 15 could be detected under these conditions (as a precipitate), suggesting that thiyl radical formation is possible under the reaction conditions, albeit as a minor reaction pathway.¶ Likewise, while RSOH 6 was not detected in D₂O/THF, a miniscule amount of 10 could be detected in CD₃OD/D₂O (Fig. 4B).

The drastically different product distribution arising from the reaction of 1 (and 9) with ¹O₂ compared to the precedent³² suggests that nucleophilic substitution on an oxidized sulfur intermediate is responsible for disulfide formation from reactions of thiols and ¹O₂, in general. While at first glance it would seem that RSOH is that intermediate, and although we found that RSOH 10 was readily photooxidized to RSO₂H 11 (see ESI Fig. S7†), it is difficult to provide a compelling mechanistic rationale for the direct formation of RSOH from RSH and ¹O₂. Instead, we posit that a sulfenyl hydroperoxide (RSOOH) is the immediate product, and when the reaction with a thiolate cannot take place at sulfur to give a disulfide (i.e. for 1 and 9), the reaction takes place at oxygen to give two equivalents of RSOH. Alternatively, RSOOH can rearrange to RSO₂H. If formed, RSOH can react with ¹O₂ to afford a sulfinyl hydroperoxide (RS(O)OOH) which can oxidize another equivalent of RSOH to yield two equivalents of RSO₂H (Fig. 4C).

Formation of RSOO⁻ from thiolate (RS⁻) and ¹O₂ is predicted by CBS-QB3 (ref. 33) to be highly exergonic (ΔG° = −25.1 kcal mol⁻¹ for t-BuS⁻ as a model thiolate). No TS could be computed for the nucleophilic addition pathway, presumably due to the use of a single determinant wavefunction approach. A TS could, however, be computed for the addition of RSH to ¹O₂, which was associated with a barrier of ΔG^‡ = 14.7 kcal mol⁻¹ (Fig. 4D). This may be taken as an upper bound for the thiolate reaction since the barrier to addition of the better nucleophile is sure to be lower. Another possibility involves electron transfer from RS⁻ to ¹O₂ to form a radical pair that can combine or, if cage escape occurs, oxygenation of the thiyl radical to yield a thiylperoxyl radical (RSOO˙) that could be subsequently reduced via either H-atom transfer or electron transfer. H-atom transfer from a thiol to RSOO˙ is predicted to be exergonic (ΔG° = −4.1 kcal mol⁻¹) and proceed via an accessible barrier (ΔG^‡ = 18.8 kcal mol⁻¹). However, our calculations indicate that the highly exergonic (ΔG° = −76.8 kcal mol⁻¹) rearrangement of RSOO˙ to a sulfonyl radical is much more likely. Earlier calculations had suggested this would require traversing a prohibitively high barrier (ΔG^‡ = 45 kcal mol⁻¹),³⁴ but we have found this barrier to be significantly lower (ΔG^‡ = 13.7 kcal mol⁻¹) using the higher accuracy CBS-QB3 methodology. Regardless, the lack of disulfide formed in our experiments argues that thiyl radicals are not key intermediates.

The sulfinate observed when disulfide formation is suppressed could also arise from concerted rearrangement of RSOOH – a highly exergonic reaction (ΔG° = −60.6 kcal mol⁻¹) with an easily accessible barrier (ΔG^‡ = 13.6 kcal mol⁻¹). Again, this barrier is likely an upper bound given that the H-bond between the migrating OH group and an explicit solvent molecule will become stronger in the TS (this cannot be captured in the polarizable solvent continuum employed in the calculation). Alternatively, a step-wise rearrangement can be envisioned. The calculated O–O BDE of RSOOH was found to be a meagre 10.2 kcal mol⁻¹, driven by the stability of the sulfinyl radical, suggesting O–O bond homolysis followed by in-cage recombination to form a sulfinic acid is possible. Earlier spin-trapping studies during thiol photooxygenation suggested the formation of ˙OH, presumably due to cage-escape following RSOOH thermolysis (Fig. 4E).³⁰ The lack of significant signal suppression in our NMR experiments, suggests that homolysis does not occur to a significant extent and that nucleophilic chemistry predominates. Regardless, the predicted rapid rearrangement chemistry presumably prevents the direct observation of RSOOH on the timescale of the NMR experiments.

The implication of RSOOH in disulfide formation from thiols exposed to ¹O₂ compels consideration that RSOOH could also be intermediates in disulfide formation in the radical-mediated oxidation (autoxidation) of thiols, given the rapid reaction of thiyl radicals with ³O₂ (∼10⁸ M⁻¹ s⁻¹).³⁵ Indeed, sulfur oxyacids are major products under some conditions,^36,37 and may derive from RSOOH as a common intermediate to disulfide formation. However, as indicated above, given that the (unimolecular) rearrangement of RSOO˙ to is predicted to proceed with a smaller barrier than the (biomolecular) H-atom transfer to RSOO˙ from RSH (13.7 kcal mol⁻¹vs. 18.8 kcal mol⁻¹, respectively), the formation of RSOOH under radical conditions is unlikely. Instead, the partitioning of RS˙ and RSOO˙ between thiyl dimerization and thiylperoxyl rearrangement pathways is likely key in determining whether disulfides or sulfur oxyacids are the major products.

2.5. Other applications of the new scaffold

Since comparably very little remains known of RSOH reactivity (relative to RSH, RSO₂H and RSO₃H), we surmised that 6 may prove generally useful as a mechanistic probe. It has been shown previously that ascorbate acts as the cofactor for 1-Cys peroxiredoxins (Prx), and it was proposed to reduce the RSOH of 1-Cys Prx back to its active thiol form after oxidation.³⁸ However, other protein sulfenic acids are resistant to reduction by ascorbate.³⁹ Previously employed indirect bifunctional assays estimated rate constants ranging from 2.2 to 4 × 10² M⁻¹ s⁻¹, depending on the protein RSOH. To provide additional insight to this reactivity, we incubated 6 with 3 eq. of ascorbate|| in pD 7.8 D₂O/THF (3 : 1) and monitored the reaction by ¹⁹F NMR (Fig. 5A). No reaction was observed over 72 hours, suggesting an amino acid sidechain in the Prx active site is required for activation of the sulfenic acid to reduction (the conjugate acid of an essential histidine residue is a possibility), or that a more sophisticated redox cascade is at play.

Fig. 5 (A) 1-Cys peroxiredoxins (Prx) reduce hydrogen peroxide and hydroperoxides by nucleophilic substitution by the cysteine thiol(ate). The resultant sulfenic acids are converted back to the starting thiol by ascorbate to complete the catalytic cycle. The lack of reactivity of RSOH 6 to ascorbate implies that active site residues are essential to enzyme turnover. (B) Lipid peroxidation is auto-initiated in the presence of Fe²⁺ due to reductive cleavage of the O–O bonds in product hydroperoxides, which yields initiating radicals. No reaction is observed for RSOH 10, implying that the process is less favourable for RSOH than it is for the corresponding ROOH.

Labile iron(II) is often invoked as key to the auto-initiation of cellular (phospho)lipid peroxidation and associated ferroptosis due to its ability to reduce lipid hydroperoxides to form initiating lipid alkoxyl radicals.^40,41 Since thiyl radicals are significantly more thermodynamically stable than alkoxyl radicals (i.e. t-BuSH and t-BuOH BDEs are 86 and 106 kcal mol⁻¹, respectively),⁴² we anticipated that RSOH would readily undergo the same reaction (Fig. 5B). To our great surprise, 10 did not react with Fe(II) (or Fe(III) in the presence of ascorbic acid) in methanol. At first glance, we surmised that this may have to do with some impairment of the reaction by the triptycene backbone. However, CBS-QB3 calculations predict that the S–O bond in a model (t-butyl) RSOH is 27 kcal mol⁻¹ stronger than the O–O bond in the corresponding hydroperoxide – despite the 20 kcal mol⁻¹ difference in the stabilities of the product thiyl and alkoxyl radicals. The significant stabilization of RSOH relative to ROOH can be rationalized by lesser lone pair/lone pair repulsion in RSOH due to the longer S–O bond and more diffuse S lone pairs and also greater lone-pair/ hyperconjugation in RSOH compared to ROOH. Based on these calculated thermodynamics and the lack of reactivity of 6 and ascorbate with or without added iron, it is clear that reductive heterolysis of RSOH is a highly unfavourable process in instances where there is no acid available to activate the RSOH.

3 Conclusions

A newly synthesized 10-fluorotriptycene thiol enables visualization of transient reactive intermediates (e.g., RSOH and RSCl) and can be used to directly observe transformations of thiols and sulfenic acids in aqueous buffer by ¹⁹F NMR. Reactions of thiols and sulfenic acids with H₂O₂ were found to proceed with kinetics that were slightly faster than in methanol, largely due to more favourable ionization to the reactive conjugate bases. Reaction of thiol with HOCl did not yield a sulfenic acid, but presumably a sulfenyl chloride. The sulfenyl chloride could be independently prepared by reaction of the thiol with N-chlorosuccinimide and was found to rapidly yield disulfide when treated with thiol(ate) in buffer – attributed to dimerization of thiyl radicals formed upon electron transfer from thiolate to sulfenyl chloride. Product distributions of the reaction of thiols with ¹O₂ enable us to propose sulfenyl and sulfinyl hydroperoxides as intermediates en route to disulfides, sulfinic and sulfonic acids. Through product analyses and CBS-QB3 calculated reaction energetics, the myriad of one- and two-electron processes that these short-lived intermediates undergo are herein summarized. A lack of reactivity of RSOH 6 and 10 to one-electron reduction strongly supports that reductive heterolysis of RSOH does not occur to a significant extent without enzyme (acid) catalysis and highlights the energetic contribution of lone pair repulsion in hydroperoxides to their relative instability. We anticipate that this scaffold will be a useful tool compound for mechanistic investigations of thiol and sulfenic acid reactivity.

4 Experimental section

4.1. General

Reagents and solvents were obtained from commercial suppliers and used as received, unless indicated otherwise. Hydrogen peroxide concentration was determined by manganometric titration. Sodium hypochlorite concentration was determined by iodometry. ¹H, ¹³C and ¹⁹F NMR spectra were recorded on a Bruker AVANCE spectrometer at the specified frequency. ¹⁹F NMR spectra in D₂O were referenced using 2,2,2-trifluoroethanol (TFE) (δ −77.0 ppm) as an internal standard, ¹⁹F NMR spectra in CDCl₃, CD₃OD, and C₆D₆ were referenced to trifluorotoluene (δ −63.72 ppm). Product ratios were obtained through integration of their ¹⁹F NMR signals. For quantification, ¹⁹F NMR spectra were phased and baseline corrected. For NMR characterization of isomeric mixtures, the signals corresponding to each isomer were assigned based on those observed for HPLC-purified isomer 1a, and the corresponding signals observed when it was oxidized with H₂O₂. Preparative HPLC was carried on a Waters 2996 HPLC system, using a preparative Waters XBridge Prep C18 5 μm (19 × 150 mm) column. Column chromatography was carried out with 40–63 μm, 230–400 mesh silica gel purchased from SiliCycle. All calculations were carried out using the Gaussian 09 quantum chemistry package using the CBS-QB3 complete basis set approach.³³

4.2. Synthesis

4.2.1 10-Fluoro-9,10-[1′,2′]benzenoanthracen-9(10H)-yl(methyl)sulfane (3). Warning: a significant amount of methanethiol is produced in this reaction. To an oven-dried flask charged with a solution of 9-bromo-10-fluorotriptycene¹⁷ (3.0 g, 8.57 mmol) in neat THF (40.0 mL) and cooled to −78 °C, was slowly added nBuLi (4.12 mL, 10.28 mmol) as a 2.5 M solution in hexanes. The reaction was stirred at −78 °C for 30 minutes, after which methyl disulfide (0.91 mL, 10.28 mmol) was added and the suspension was left to warm up to room temperature, upon which it became a homogeneous solution. After 1 h, the reaction was quenched with a saturated solution of NH₄Cl. The crude product was extracted using EtOAc (3 × 50 mL), dried over MgSO₄ and concentrated in vacuo. The residue was then purified by column chromatography using hexanes as the eluent to yield 3 (2.51 g, 92%) as a colourless solid: ¹H NMR (400 MHz, CDCl₃) δ 7.65–7.59 (m, 6H), δ 7.17–7.09 (m, 6H), δ 2.58 (s, 3H). ¹³C NMR (101 MHz, CDCl₃) δ 144.2, 144.0, 141.7, 141.6, 125.7, 125.6, 122.5, 118.7, 97.6, 60.5, 14.1. HRMS (EI) m/z calculated for [M] C₂₁H₁₅FS 318.0879, found 318.0866.

4.2.2 Dibromo-10-fluoro-9,10-[1′,2′]benzenoanthracen-9(10H)-yl(methyl)sulfane (4). An oven-dried Schlenk tube was charged with 3 (1.74 g, 5.47 mmol), TpSMe¹⁸ (82 mg, 0.274 mmol) and DCE (25 mL). Under a flow of nitrogen, N-bromosuccinimide (3.51 g, 19.69 mmol) and AgSbF₆ (93 mg, 0.274 mmol) were both added as solids. The reaction was stirred for 48 h at 50 °C before being poured into a saturated NaHCO₃ solution (50 mL). The product was extracted using CHCl₃ (3 × 30 mL), concentrated in vacuo, and dried over MgSO₄. The residue was purified by column chromatography using hexanes as the eluent to give a mixture of dibrominated isomers as an off-white foam (1.58 g, 61% total yield). The monobrominated derivative was also isolated.
4.2.2.1 2,7-Dibromo-10-fluoro-9,10-[1′,2′]benzenoanthracen-9(10H)-yl(methyl)sulfane (4a). 33% yield. ¹H NMR (600 MHz, CDCl₃) δ 7.74–7.71 (m, 2H), δ 7.61–7.57 (m, 2H), δ 7.47–7.43 (m, 2H), δ 7.31–7.24 (m, 2H), δ 7.19–7.13 (m, 2H), δ 2.56 (s, 3H) ¹³C NMR (151 MHz, CDCl₃) δ 145.9, 143.6, 143.2, 142.9, 142.8, 142.7, 140.7, 140.5, 128.9, 126.2, 126.1, 122.8, 122.4, 120.5, 119.0, 97.8, 59.9, 14.1. HRMS (EI) m/z calculated for [M] C₂₁H₁₃Br₂FS 473.9087, found 473.9089.
4.2.2.2 2,6-Dibromo-10-fluoro-9,10-[1′,2′]benzenoanthracen-9(10H)-yl(methyl)sulfane (4b). 28% yield. ¹H NMR (600 MHz, CDCl₃) δ 7.74–7.71 (m, 2H), δ 7.61–7.57 (m, 2H), δ 7.47–7.43 (m, 2H), δ 7.31–7.24 (m, 2H), δ 7.19–7.13 (m, 2H), δ 2.54 (s, 3H) ¹³C NMR (151 MHz, CDCl₃) δ 145.7, 143.4, 143.1, 142.6, 140.7, 140.5, 128.9, 126.2, 126.1, 124.4, 122.8, 122.4, 120.5, 119.9, 118.9, 97.0, 59.9, 14.0. HRMS (EI) m/z calculated for [M] C₂₁H₁₃Br₂FS 473.9087, found 473.9089.

4.2.3 Di(tetraethylene glycol)-10-fluoro-9,10-[1′,2′]benzenoanthracen-9(10H)-yl(methyl)sulfane (5). To an oven-dried Schlenk tube was added a mixture of the isomers 4 (1.10 g, 2.32 mmol) suspended in tetraethylene glycol (20.0 mL). Copper(II) acetate hydrate (0.91 g, 4.58 mmol) and potassium carbonate (3.20 g, 23.15 mmol) were then both added as solids. The tube was sealed, degassed, and refilled with nitrogen. The blue solution was heated at 150 °C for 5 days, during which the reaction became brown. Once the reaction mixture was cooled to ambient temperature, it was poured into distilled water (100 mL) and extracted (3 × 25 mL) with CH₂Cl₂. The organic layer was washed with water (3 × 100 mL) to remove excess TEG, dried over MgSO₄, and concentrated in vacuo to obtain the crude product as a dark-orange oil, which was then purified by column chromatography (gradient 1–3% MeOH/CH₂Cl₂) to obtain the mixture of isomers 5 (5a : 5b = 57 : 43) as a viscous yellow oil (1.27 g, 1.81 mmol, 78% total yield).
4.2.3.1 2,7-Di(tetraethylene glycol)-10-fluoro-9,10-[1′,2′]benzenoanthracen-9(10H)-yl(methyl)sulfane (5a). ¹H NMR (600 MHz, CDCl₃) δ 7.58–7.52 (m, 2H), 7.46–7.40 (m, 2H), 7.43 (d, J = 8.2 Hz, 2H), 7.23–7.16 (m, 2H), 7.14–7.06 (m, 2H), 6.64–6.57 (m, 2H), 4.08 (t, J = 4.8 Hz, 4H), 3.80 (t, J = 4.8 Hz, 4H), 3.73–3.61 (m, 20H), 3.57 (t, J = 4.5 Hz, 4H), 2.56 (s, 3H). ¹³C NMR (151 MHz, CDCl₃) δ 156.9, 143.3, 141.5, 137.0, 136.1, 125.7, 125.5, 122.5, 122.2, 119.3, 118.2, 111.3, 110.8, 110.2, 106.2, 97.0, 72.5, 70.8, 70.6, 70.3, 69.7, 67.8, 61.7, 60.6, 60.0, 14.2. HRMS (ESI+) m/z calculated for [M + Na]⁺ C₃₆H₄₅FNaO₁₀S 725.2772, found 725.2767.
4.2.3.2 2,6-Di(tetraethylene glycol)-10-fluoro-9,10-[1′,2′]benzenoanthracen-9(10H)-yl(methyl)sulfane (5b). ¹H NMR (600 MHz, CDCl₃) δ 7.58–7.52 (m, 2H), 7.46–7.40 (m, 2H), 7.43 (d, J = 8.2 Hz, 2H), 7.23–7.16 (m, 2H), 7.14–7.06 (m, 2H), 6.64–6.57 (m, 2H), 4.08 (t, J = 4.8 Hz, 4H), 3.80 (t, J = 4.8 Hz, 4H), 3.73–3.61 (m, 20H), 3.57 (t, J = 4.5 Hz, 4H), 2.54 (s, 3H). ¹³C NMR (151 MHz, CDCl₃) δ 157.0, 156.9, 143.3, 137.0, 125.7, 125.5, 123.6, 122.5, 122.2, 119.5, 119.3, 118.3, 111.3, 110.8, 110.0, 97.0, 72.5, 70.8, 70.6, 70.3, 69.7, 67.8, 61.7, 60.6, 60.0, 14.2. HRMS (ESI+) m/z calculated for [M + Na]⁺ C₃₆H₄₅FNaO₁₀S 725.2772, found 725.2767.

4.2.4 Di(tetraethylene glycol)-10-fluoro-9,10-[1′,2′]benzenoanthracen-9(10H)-ylmercaptan (1). In a sealable tube, methyl sulfide 5 (349 mg, 0.497 mmol) was dissolved in anhydrous DMPU (2.0 mL), and the solution was purged with N₂ for 2 minutes before freshly prepared lithium ethylmercaptide⁴³ (169 mg, 2.485 mmol) was quickly added and the tube sealed. The reaction was heated at 120 °C for 2 days. Once cooled to room temperature, it was poured into 0.1 M aqueous HCl (100 mL) and extracted with CH₂Cl₂ (3 × 20 mL). The organic fraction was then washed with 0.5 M NaOH (3 × 75 mL), the aqueous washings were collected and acidified to pH ∼2 using 12 M HCl, and extracted with CH₂Cl₂ (3 × 50 mL), dried over MgSO₄, and concentrated in vacuo to give the crude product as a red oil that was purified by column chromatography (gradient elution, 2–4% MeOH/CHCl₃) to give the mixture of thiols 1 (1a : 1b = 57 : 43) as a viscous pale-yellow oil (174 mg, 51% total yield). The 1a isomer could be separated by reverse phase HPLC using 3 : 1 MeOH : H₂O as the mobile phase.
4.2.4.1 2,7-Di(tetraethylene glycol)-10-fluoro-9,10-[1′,2′]benzenoanthracen-9(10H)-ylmercaptan (1a). δ 156.9 (s), 144.6 (d, J = 6.0 Hz), 143.8 (d, J = 20 Hz), 136.2 (d, J = 20 Hz), 126.1 (s), 125.5 (s), 121.4 (d, J = 2 Hz), 119.1 (d, J = 6.0 Hz), 118.0 (d, J = 6 Hz), 110.4 (s), 110.2 (s), 97.5 (d, J = 205 Hz), 72.5 (s), 70.8 (s), 70.6 (s), 70.3 (s), 69.7 (s), 67.8 (s), 61.7 (s), 57.5 (d, J = 2 Hz). ¹⁹F NMR (471 MHz, D₂O/THF) δ −207.1. HRMS (ESI+) m/z calculated for [M + Na]⁺ C₃₆H₄₅FNaO₁₀S 711.2632, found 711.2615.
4.2.4.2 2,6-Di(tetraethylene glycol)-10-fluoro-9,10-[1′,2′]benzenoanthracen-9(10H)-ylmercaptan (1b). ¹H NMR (600 MHz, CDCl₃) δ 7.63 (m, 1H), 7.57–7.49 (m, 2H), 7.43 (m, 1H), 7.31 (m, 1H), 7.18 (d, J = 2.4 Hz, 1H), 7.17–7.04 (m, 2H), 6.62 (dd, J = 8.4, 2.4 Hz, 1H), 6.58 (dd, J = 8.4, 2.4 Hz, 1H), 4.09 (t, J = 4.8 Hz, 4H), 3.80 (t, J = 4.8 Hz, 4H), 3.73–3.61 (m, 20H), 3.57 (t, J = 4.5 Hz, 4H), 2.42 (s, 1H). ¹³C NMR (151 MHz, CDCl₃) δ 157.1 (s), 157.0 (s), 145.0 (d, J = 20 Hz), 143.5 (d, J = 20 Hz), 143.2 (d, J = 6 Hz), 135.4 (d, J = 20 Hz), 125.8 (s), 125.7 (s), 122.5 (d, J = 2 Hz), 121.1 (d, J = 2 Hz), 119.3 (d, J = 2 Hz), 118.2 (d, J = 2 Hz), 110.6 (s), 110.5 (s), 110.3 (s), 110.0 (d, J = 2 Hz), 97.5 (d, J = 205 Hz), 72.5 (s), 70.8 (s), 70.6 (s), 70.3 (s), 70.3 (s), 69.7 (s), 67.8 (s), 61.7 (s), 57.5 (d, J = 2 Hz). ¹⁹F NMR (471 MHz, D₂O/THF) δ −206.3. HRMS (ESI+) m/z calculated for [M + Na]⁺ C₃₆H₄₅FNaO₁₀S 711.2632, found 711.2615.

4.2.5 Di(tetraethylene glycol)-10-fluoro-9,10-[1′,2′]benzenoanthracen-9(10H)-ylsulfenic acid (6a) and potassium di(tetraethylene glycol)-10-fluoro-9,10-[1′,2′]benzenoanthracen-9(10H)-ylsulfinate (7). Thiols 1 (60 mg, 0.087 mmol, as the isomeric mixture of 1a : 1b 57 : 43) were dissolved in THF (1 mL) and added to a solution of dibasic potassium phosphate (30.4 mg, 0.194 mmol) in HPLC-grade H₂O (900 μL). The mixture was thoroughly mixed to ensure homogeneity before 100 μL of H₂O₂ as a 1.94 M solution in D₂O/H₂O was slowly added while stirring. Progress of the reaction was monitored by ¹⁹F NMR. Upon complete consumption of the starting material the reaction was concentrated in vacuo until almost no water remained. The residue was purified by column chromatography (gradient elution 4–10% MeOH/CHCl₃) to obtain the isomeric sulfenic acids 6 (28.2 mg, 46% total yield, 6a : 6b = 58 : 42) and sulfinates 7 (19.1 mg, 22% total yield, 7a : 7b = 57 : 43). Isomerically pure 6a was obtained using the same protocol starting from thiol 1a. To obtain isomerically pure 7a, sulfenic acid 6a (5.8 mg, 0.00823 mmol) was dissolved in a 10 mM solution of NaOH in CD₃OD (0.75 mL) in an NMR tube. 25 μL of H₂O₂ as a 300 mM solution in CD₃OD was then added in 5 μL portions until quantitative conversion of 6a to 7a was observed by ¹⁹F NMR.
4.2.5.1 2,7-Di(tetraethylene glycol)-10-fluoro-9,10-[1′,2′]benzenoanthracen-9(10H)-ylsulfenic acid (6a). ¹H NMR (600 MHz, CD₃OD) δ 7.53–7.48 (m, 2H), 7.41–7.38 (d, J = 8.2 Hz, 2H), 7.23 (t, J = 2.1 Hz, 2H), 7.14–7.05 (m, 2H), 6.67 (dd, J = 8.2, 2.1 Hz, 2H), 4.08 (m, 4H), 3.77 (m, 4H), 3.67–3.63 (m, 4H), 3.63–3.57 (m, 12H), 3.56–3.52 (m, 4H), 3.50–3.46 (m, 4H). ¹³C NMR (151 MHz, CD₃OD) δ 158.4 (s), 146.2 (d, J = 20 Hz), 144.5 (d, J = 6 Hz), 142.7 (d, J = 6 Hz), 138.3 (d, J = 20 Hz), 128.8 (s), 126.6 (d, J = 6 Hz), 123.3 (d, J = 2.5 Hz), 119.8 (d, J = 6 Hz), 118.8 (d, J = 6 Hz), 112.2 (d, J = 2.5 Hz), 111.4 (s), 111.4 (s), 99.0 (d, J = 200 Hz), 73.6 (s), 71.7 (s), 71.6 (s), 71.5 (s), 71.3 (s), 71.3 (s), 70.8 (s), 68.9 (s), 64.5 (d, J = 2.5 Hz), 62.2 (s). ¹⁹F NMR (471 MHz, D₂O/THF) δ −206.65. HRMS (ESI+) m/z calculated for [M + Na]⁺ C₃₆H₄₅FNaO₁₁S 727.2564, found 727.2592.
4.2.5.2 2,6-Di(tetraethylene glycol)-10-fluoro-9,10-[1′,2′]benzenoanthracen-9(10H)-ylsulfenic acid (6b). ¹H NMR (600 MHz, CD₃OD) δ 7.53–7.47 (m, 2H), 7.50–7.43 (m, 2H), 7.31–7.27 (m, 1H), 7.17–7.05 (m, 3H), 6.68 (dd, J = 8.2, 2.4 Hz, 1H), 6.62 (dd, J = 8.2, 2.4 Hz, 1H) 4.16 (m, 4H), 3.85 (m, 4H), 3.74–3.57 (m, 20H), 3.54 (t, J = 4.5 Hz, 4H). ¹³C NMR (151 MHz, CD₃OD) δ 158.6 (s), 158.3 (s), 147.5 (d, J = 20 Hz), 145.7 (d, J = 20 Hz), 145.0 (d, J = 6 Hz), 143.2 (d, J = 6 Hz), 137.8 (d, J = 6 Hz), 135.3 (d, J = 20 Hz), 126.8 (s), 126.5 (s), 125.4 (s), 124.5 (d, J = 2.5 Hz), 123.1 (d, J = 2.5 Hz), 120.0 (d, J = 6 Hz), 119.0 (d, J = 6 Hz), 111.9 (d, J = 2.5 Hz), 111.7 (s), 111.3 (s), 99.0 (d, J = 200 Hz), 74.5 (s), 72.6 (s), 72.4 (s), 72.2 (s), 71.7 (s), 69.8 (s), 64.1 (d, J = 2.5 Hz), 63.1 (s). ¹⁹F NMR (471 MHz, D₂O/THF) δ −205.95. HRMS (ESI+) m/z calculated for [M + Na]⁺ C₃₆H₄₅FNaO₁₁S 727.2564, found 727.2592.
4.2.5.3 2,7-Di(tetraethylene glycol)-10-fluoro-9,10-[1′,2′]benzenoanthracen-9(10H)-ylsulfinate (7a). ¹H NMR (500 MHz, CD₃OD) δ 8.22 (br, 2H), 7.96 (br, 1H), 7.46 (m, 1H), 7.37 (d, J = 8.2 Hz, 2H), 7.11–7.04 (m, 2H), 6.67 (dd, J = 8.2 Hz, 2.1 Hz, 2H), 4.15 (m, 4H), 3.81 (m, 4H), 3.71–3.56 (m, 20H), 3.54–3.49 (m, 4H). ¹³C NMR (126 MHz, CD₃OD) δ 157.3 (s), 157.2 (s), 149.1 (d, J = 20 Hz), 147.7 (d, J = 6 Hz), 143.8 (d, J = 6 Hz), 142.6 (d, J = 6 Hz), 140.2 (d, J = 20 Hz), 128.8 (s), 126.0 (d, J = 20 Hz), 125.4 (d, J = 2.5 Hz), 119.7 (d, J = 6 Hz), 118.6 (d, J = 6 Hz), 113.9 (d, J = 2.5 Hz), 111.1 (s), 99.1 (d, J = 200 Hz), 72.3 (d, J = 2 Hz), 72.1 (s), 70.3 (s), 70.2 (s), 70.1 (s), 70.0 (s), 69.3 (s), 67.5 (s), 60.6 (s). ¹⁹F NMR (471 MHz, D₂O/THF) δ −206.1. HRMS (ESI−) m/z calculated for [M–H]⁻ C₃₆H₄₄FO₁₂S 719.2538, found 719.2526.
4.2.5.4 2,6-Di(tetraethylene glycol)-10-fluoro-9,10-[1′,2′]benzenoanthracen-9(10H)-ylsulfinate (7b). ¹H NMR (500 MHz, CD₃OD) δ 8.50 (br, 1H), 7.97 (br, 1H), 7.47 (m, 1H), 7.36 (m, 1H), 7.14–7.04 (m, 3H), 6.71–6.62 (m, 2H), 4.15 (m, 4H), 3.81 (m, 4H), 3.71–3.56 (m, 20H), 3.54–3.49 (m, 4H). ¹³C NMR (126 MHz, CD₃OD) δ 157.5 (s), 157.4 (s), 145.1 (d, J = 20 Hz), 144.6 (d, J = 20 Hz), 142.0 (d, J = 6 Hz), 141.6 (d, J = 6 Hz), 137.1 (d, J = 20 Hz), 126.1 (s), 125.1 (s), 125.0 (d, J = 2.5 Hz), 119.4 (d, J = 6 Hz), 118.0 (d, J = 6 Hz), 117.0 (d, J = 6 Hz), 116.8 (d, J = 6 Hz), 114.3 (d, J = 2.5 Hz), 113.1 (d, J = 2.5 Hz), 110.5 (s), 110.3 (s), 99.9 (d, J = 200 Hz), 72.4 (d, J = 2.5 Hz), 72.1 (s), 70.3 (s), 70.2 (s), 70.1 (s), 70.0 (s), 69.3 (s), 67.5 (s), 60.6 (s). ¹⁹F NMR (471 MHz, D₂O/THF) δ −205.4. HRMS (ESI−) m/z calculated for [M–H]⁻ C₃₆H₄₄FO₁₂S 719.2538, found 719.2526.

4.2.6 Potassium di(tetraethylene glycol)-10-fluoro-9,10-[1′,2′]benzenoanthracen-9(10H)-ylsulfonate (8). Thiols 1 (30 mg, 0.044 mmol, as a 53 : 47 isomeric mixture of 1a : 1b) or purified 1a were dissolved in THF (1 mL) and added to a solution of dibasic potassium phosphate (15 mg, 0.086 mmol) in HPLC-grade H₂O (900 μL). 114 μL of H₂O₂ as a 1.94 M solution in D₂O/H₂O was added all at once. After stirring overnight, the reaction was concentrated in vacuo until almost no water remained. The residue was purified by column chromatography (10% MeOH/CHCl₃) to obtain the isomeric sulfonates 8 (30 mg, quantitative yield, 8a : 8b = 53 : 47).
4.2.6.1 2,7-Di(tetraethylene glycol)-10-fluoro-9,10-[1′,2′]benzenoanthracen-9(10H)-ylsulfonate (8a). ¹H NMR (500 MHz, CD₃OD) δ 8.23 (m, 1H), 7.91 (t, J = 2.1 Hz, 2H), 7.51–7.45 (m, 1H), 7.41–7.35 (m, 2H), 7.14–7.04 (m, 2H), 6.71–6.63 (m, 2H), 4.15 (m, 4H), 3.81 (m, 4H), 3.71–3.56 (m, 20H), 3.54–3.49 (m, 4H). ¹³C NMR (126 MHz, CD₃OD) δ 156.5 (s), 145.1 (d, J = 20 Hz), 141.5 (d, J = 6 Hz), 140.0 (d, J = 6 Hz), 137.5 (d, J = 20 Hz), 125.1 (s), 124.9 (d, J = 2 Hz), 124.8 (s), 117.9 (d, J = 6 Hz), 116.9 (d, J = 6 Hz), 113.0 (d, J = 2 Hz), 110.4 (s), 97.5 (d, J = 200 Hz), 73.6 (d, J = 2 Hz), 73.4 (s), 71.6 (s), 71.5 (s), 71.4 (s), 71.3 (s), 71.2 (s), 68.9 (s), 62.2 (s), 62.0 (s). ¹⁹F NMR (471 MHz, D₂O/THF) δ −205.9. HRMS (ESI−) m/z calculated for [M–H]⁻ C₃₆H₄₄FO₁₃S 735.2503, found 735.2491.
4.2.6.2 2,6-Di(tetraethylene glycol)-10-fluoro-9,10-[1′,2′]benzenoanthracen-9(10H)-ylsulfonate (8b). ¹H NMR (500 MHz, CD₃OD) δ 8.23 (m, 1H), 8.12 (dd, J = 8.4, 1.4 Hz, 1H), 7.89 (t, J = 2.1 Hz, 1H), 7.51–7.45 (m, 1H), 7.41–7.35 (m, 1H), 7.14–7.04 (m, 3H), 6.71–6.63 (m, 2H), 4.15 (m, 4H), 3.81 (m, 4H), 3.71–3.56 (m, 20H), 3.54–3.49 (m, 4H). ¹³C NMR (126 MHz, CD₃OD) δ 156.7 (s), 156.6 (s), 146.4 (d, J = 20 Hz), 144.5 (d, J = 20 Hz), 142.0 (d, J = 6 Hz), 140.0 (d, J = 6 Hz), 137.0 (d, J = 20 Hz), 132.5 (d, J = 6 Hz), 126.2 (d, J = 2 Hz), 125.0 (s), 124.9 (d, J = 2 Hz), 124.7 (d, J = 2 Hz), 118.0 (d, J = 6 Hz), 117.0 (d, J = 6 Hz), 112.7 (d, J = 2 Hz), 110.4 (s), 110.0 (s), 97.5 (d, J = 200 Hz), 73.6 (d, J = 2 Hz), 73.4 (s), 71.6 (s), 71.5 (s), 71.4 (s), 71.3 (s), 71.2 (s), 68.9 (s), 62.2 (s), 62.0 (s). ¹⁹F NMR (471 MHz, D₂O/THF) δ −205.25. HRMS (ESI−) m/z calculated for [M–H]⁻ C₃₆H₄₄FO₁₃S 735.2503, found 735.2487.

4.2.7 2,7-Di(tetraethylene glycol)-10-fluoro-9,10-[1′,2′]benzenoanthracen-9(10H)-yl disulfide (13a). In an NMR tube, thiol 1a (8.2 mg, 11.9 μmol) was dissolved in THF (175 μL) containing 240 mM of TFE (internal standard). Phosphate-buffered saline (490 μL of a 50 mM pD 7.8 solution in D₂O) was then added. Upon reaching homogeneity, 17 μL of NaOCl (as a 700 mM solution in D₂O/H₂O) was added portion-wise, monitoring the consumption of 1a. The tube was mixed, and the reaction was complete upon acquiring the next set of scans (ca. 45 s). The contents of the NMR tube were diluted with distilled H₂O (10 mL) and extracted with CHCl₃ (3 × 5 mL), dried over MgSO_4, and concentrated in vacuo to give 13a (8.2 mg, 11.9 μmol). ¹H NMR (600 MHz, CDCl₃) δ 8.11 (d, J = 7.5 Hz, 2H), 7.55 (d, 2H), 7.39 (d, J = 8.2 Hz, 4H), 7.33 (t, J = 2.1 Hz, 4H), 7.13 (td, J = 7.5 Hz, 1.3 Hz, 2H), 7.04 (td, J = 7.5 Hz, 1.3 Hz, 2H), 6.56 (d, J = 8.2 Hz, 2.1 Hz, 4H), 4.15 (m, 8H), 3.81 (m, 8H), 3.71–3.56 (m, 40H), 3.54–3.49 (m, 4H), 2.78 (br s, exchangeable Hs). ¹³C NMR (151 MHz, CDCl₃) δ 156.4 (s), 144.7 (d, J = 20 Hz), 142.8 (d, J = 6 Hz), 141.5 (s), 136.3 (d, J = 20 Hz), 126.1 (s), 125.3 (s), 123.1 (s), 119.0 (d, J = 2.5 Hz), 118.1 (d, J = 2.5 Hz), 118.1 (d, J = 2.5 Hz), 111.7 (s), 111.4 (s), 97.2 (d, J = 200 Hz), 72.5 (s), 70.6 (s), 70.5 (s), 70.3 (s), 69.5 (s), 67.0 (s), 62.8 (d, J = 2.5 Hz), 61.7 (s). ¹⁹F NMR (471 MHz, D₂O/THF) δ −205.9. HRMS (ESI+) m/z calculated for [M + Na]⁺ C₇₂H₈₈F₂O₂₀NaS₂ 1397.5176, found 1397.5193.

4.2.8 2,7-Di(tetraethylene glycol)-10-fluoro-9,10-[1′,2′]benzenoanthracen-9(10H)-ylsulfenyl chloride (14a). In an NMR tube, thiol 1a (6.9 mg, 10.0 μmol) was dissolved in CDCl₃ (700 μL) containing TFE (internal standard). N-Chlorosuccinimide as a 200 mM solution in CDCl₃ was added portion-wise every 5 minutes until the starting thiol was completely consumed (ca. 50 μL, 1 eq.). The contents of the NMR tube were washed with distilled H₂O (5 mL) and extracted with CHCl₃ (3 × 3 mL), dried over MgSO₄, and concentrated in vacuo to give 14a (6.8 mg, 94% yield). ¹H NMR (500 MHz, CDCl₃) δ 7.56 (dd, J = 7.4 Hz, 1.0 Hz, 2H), 7.45 (d, J = 8.2 Hz, 2H), 7.39 (d, J = 8.2 Hz, 2H), 7.23 (br m, 2H), 7.17–7.07 (m, 2H), 6.65 (d, J = 8.2 Hz, 2.1 Hz, 2H), 4.09 (m, 4H), 3.81 (m, 4H), 3.72–3.61 (m, 20H), 3.58–3.54 (m, 4H), 2.61 (br s, exchangeable H's). ¹³C NMR (151 MHz, CD₃OD) δ 156.9 (s), 144.4 (d, J = 20 Hz), 141.3 (d, J = 6 Hz), 139.8 (d, J = 6 Hz), 136.6 (d, J = 20 Hz), 126.2 (s), 125.6 (s), 122.1 (br), 119.4 (d, J = 6 Hz), 118.4 (d, J = 6 Hz), 110.8 (s), 97.4 (d, J = 200 Hz), 72.5 (s), 70.8 (s), 70.7 (s), 70.6 (s), 70.3 (s), 69.6 (s), 67.8 (s), 63.8 (d, J = 2 Hz), 61.7 (s). ¹⁹F NMR (471 MHz, D₂O/THF) δ −206.4. HRMS (ESI+) m/z calculated for [M + Na]⁺ C₃₆H₄₄FClO₁₀S 745.2225, found 745.2249.

4.2.9 10-Fluoro-9,10-[1′,2′]benzenoanthracen-9(10H)-yl disulfide (15). Thiol 9 (ref. 17) (30 mg, 0.1 mmol) was suspended in MeOH (5 mL), potassium carbonate (45 mg, 0.3 mmol) was added as a solid, and the mixture was stirred for 10 minutes at room temperature. Iodine (50 mg, 0.2 mmol) was then added directly and the reaction was stirred for 2 hours. The reaction was then poured into an aqueous solution (50 mL) of sodium thiosulfate pentahydrate (100 mg, 0.4 mmol), and subsequently extracted with EtOAc (3 × 25 mL). The organic layer was dried over MgSO₄ and concentrated in vacuo. The residue was purified by column chromatography (10% EtOAc/hexanes) to obtain disulfide 15 (11 mg, 37%). Although conversion looked to be complete by TLC, the isolated yield was reduced due to low solubility of the compound and corresponding losses up during chromatography. ¹H NMR (600 MHz, CDCl₃) δ 7.99 (dd, J = 7.6, 1.1 Hz, 6H), 7.61 (dd, J = 7.6, 1.1 Hz, 6H), 7.11 (td, J = 7.6, 1.1 Hz, 6H), 6.94 (td, J = 7.6, 1.1 Hz, 6H). ¹³C NMR (151 MHz, CDCl₃) δ 146.1, 143.5, 143.3, 141.5, 126.0, 125.5, 123.3, 118.0, 96.6, 62.8. ¹⁹F NMR (282 MHz, C₆D₆) δ −206.6. HRMS (EI) m/z calculated for [M]⁺ C₄₀H₂₄F₂S₂ 606.1288, found 606.1287.

4.3 ¹⁹F NMR-monitored reactions

4.3.1 Buffer preparation. Mixtures of K₂HPO₄ and KH₂PO₄ were dissolved in D₂O to a final concentration of 50 mM. Either KOH or HCl was added to adjust the pH to the desired value. Observed pH (pD) was converted to pH by subtracting 0.4.¹⁹

4.3.2 Oxidation of 1 with H₂O₂. Solutions of 1 (typically 15–20 mM) in THF containing 240 mM TFE were prepared before each experiment. 175 μL of this solution was transferred to a NMR tube. 525 μL of phosphate-buffered D₂O was added to the NMR tube and the mixture was vortexed to homogeneity. An initial set of scans was acquired and then H₂O₂ was added as a 1.94 M solution in D₂O/H₂O. The NMR tube was mixed before the reaction was monitored for 15 hours. COPASI⁴⁴ was used to fit the data to a simple kinetic scheme for successive bimolecular reactions of thiol, sulfenic and sulfinic acids with H₂O₂, which yielded the observed rate constants given in the text.

4.3.3 Oxidation of 1a with HOCl. A solution of 1a (68 mM) in THF containing 240 mM TFE was prepared. An aliquot (175 μL) of this solution was transferred to an NMR tube and 490 μL of phosphate-buffered D₂O (pD 7.8) was added. An initial spectrum was recorded, after which 0.35 eq. of NaOCl (as a solution in D₂O) was added at once. The tube was mixed and the spectrum was recorded, revealing immediate quantitative conversion to the disulfide 13a.

4.3.4 Reduction of sulfenyl chloride 14a with 1a. A solution of 14a (50 mM) in THF containing 240 mM TFE was prepared. An aliquot (175 μL) of this solution was transferred to an NMR tube and 490 μL of phosphate-buffered D₂O (pD 7.8) was added. The NMR tube was left overnight under ambient conditions to rule out hydrolysis as a decomposition pathway. After 18 h an initial set of spectra was acquired and a solution of 1a (20 mM) in D₂O/THF (3 : 1) was added dropwise, while mixing, revealing stoichiometric conversion of 14a to 13a.

4.3.5 Oxidation of 1 with thermally generated ¹O₂. A 20 mM solution of 1 in THF containing 240 mM TFE was prepared before the experiment. 175 μL of this solution was transferred to an NMR tube and 525 μL of an ice-cooled freshly prepared 27 mM solution of NDPO₂ (prepared using a reported procedure⁴⁵) in phosphate-buffered D₂O (pD 7.8) was added to the NMR tube. The NMR tube was mixed and an initial set of scans at 25 °C was acquired before the NMR spectrometer was heated to 37 °C using a thermostat. The reaction was then monitored for 30 min. Due to significant deviation in chemical shift values from the conditions involving oxidation of 1 by H₂O₂, a reaction mixture obtained by reacting thiol 1 with H₂O₂ was added to the reaction mixture after completion to identify the observed products.

4.3.6 Oxidation of 9 with photochemically generated ¹O₂. In an NMR tube, 9 (6 mg, 0.02 mmol) was suspended in CD₃OD (690 mL), to this suspension was added sodium hydroxide as a 1 M solution in water (60 μL, 0.06 mmol, 3 eq.) and the mixture became homogeneous. 10 μL of a 1.5 mM aqueous solution of Rose Bengal (0.0002 mmol. 0.01 eq.) was added and the reaction mixture was cooled to ca. 4 °C. Trifluorotoluene (10 mM) was used as an internal standard for NMR quantification. Oxygen was bubbled through the solution after which the reaction vessel was irradiated with a 400 W HPS lamp. Time points were taken to analyze reaction progress. The reaction was continued to the point that 9 was fully consumed, after which the reaction was concentrated in vacuo. Following this, the crude product was concentrated in vacuo, suspended in C₆D_6, and analyzed by ¹⁹F NMR, which resulted in the observation of disulfide 14 as a trace reaction product, 11 and 12 were not observed despite being major products as they are insoluble in C₆D₆ (Fig. S5†). Products were assigned after adding authentic 12 to the NMR tube (CD₃OD). Monitoring of reaction progress is shown in Fig. 3B of the manuscript. ¹H NMR spectra of the initial and final (after 6 min. of irradiation) time points are shown in Fig. S4.†

4.3.7 Reduction of sulfenic acid 6 with ascorbate. A 5 mM solution of 6 in THF containing 10 mM TFE was prepared before the experiment. 175 μL of this solution was transferred to an NMR tube and 300 μL of phosphate-buffered pD 7.8 D₂O was added to the NMR tube before an initial set of scans was acquired. No spectroscopic changes were observed after 24 h. An additional time point was taken after 3 days, and no changes were observed.

4.3.8 Reduction of sulfenic acid 10 with ascorbate and iron. Sulfenic acid 10 (6.4 mg, 0.02 mmol) was dissolved in MeOH (8 mL). 1 mL of a freshly prepared 200 mM solution of ascorbic acid in MeOH (1 eq.) was then added, followed by addition of a 20 mM solution of Fe₂(SO₄)₃ in MeOH (1 mL, 0.2 eq.). The mixture was stirred for 24 hours, after which it was diluted with CH₂Cl₂ (50 mL), filtered through a silica plug and washed with 20% MeOH/CH₂Cl₂ (50 mL). The filtrate was concentrated in vacuo and analyzed by ¹⁹F NMR. No changes were observed, with the mass balance indicating that no reaction had occurred.

Data availability

Additional experimental procedures and reaction monitoring data, NMR spectra and computational data are available in the ESI.†

Author contributions

D. D. S. designed and executed the experiments and carried out quantum chemical calculations. D. A. P. conceived the project and contributed to experimental design. D. D. S. and D. A. P. wrote the manuscript.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by grants from the Natural Sciences and Engineering Council of Canada and the Canada Foundation for Innovation to D. A. P. Computational investigations were enabled by Compute Ontario and the Digital Research Alliance of Canada. D. D. S. thanks the Government of Ontario for support in the form of an Ontario Graduate Scholarship.

Notes and references

1. B. Morgan, D. Ezeriņa, T. N. E. Amoako, J. Riemer, M. Seedorf and T. P. Dick, Nat. Chem. Biol., 2013, 9, 119–125.
2. G. J. McBean, M. Aslan, H. R. Griffiths and R. C. Torrão, Redox Biol., 2015, 5, 186–194.
3. J.-M. Lv, S.-Q. Lü, Z.-P. Liu, J. Zhang, B.-X. Gao, Z.-Y. Yao, Y.-X. Wu, L. A. Potempa, S.-R. Ji, M. Long and Y. Wu, Sci. Rep., 2018, 8, 1494.
4. S. García-Santamarina, S. Boronat and E. Hidalgo, Biochemistry, 2014, 53, 2560–2580.
5. S. G. Rhee, Science, 2006, 312, 1882–1883.
6. Z. Wu, U. Barayeu, D. Schilling, T. P. Dick and D. A. Pratt, Curr. Opin. Chem. Biol., 2023, 76, 102353.
7. Y.-C. Chang, C.-N. Huang, C.-H. Lin, H.-C. Chang and C.-C. Wu, Proteomics, 2010, 10, 2961–2971.
8. S. Akter, L. Fu, Y. Jung, M. Lo Conte, J. R. Lawson, W. T. Lowther, R. Sun, K. Liu, J. Yang and K. S. Carroll, Nat. Chem. Biol., 2018, 14, 995–1004.
9. J. M. M. Pople and J. M. Chalker, Curr. Opin. Chem. Biol., 2021, 60, 55–65.
10. N. J. Kettenhofen and M. J. Wood, Chem. Res. Toxicol., 2010, 23, 1633–1646.
11. V. Gupta and K. S. Carroll, Biochim. Biophys. Acta, Gen. Subj., 2014, 1840, 847–875.
12. C. C. Winterbourn and D. Metodiewa, Free Radical Biol. Med., 1999, 27, 322–328.
13. N. Nakamura, J. Am. Chem. Soc., 1983, 105, 7172–7173.
14. T. Yoshimura, E. Tsukurimichi, S. Yamazaki, S. Soga, C. Shimasaki and K. Hasegawa, J. Chem. Soc. Chem. Commun., 1992, 1337–1338.
15. T. Sano, R. Masuda, S. Sase and K. Goto, Chem. Commun., 2021, 57, 2479–2482.
16. C. C. Winterbourn and M. B. Hampton, Free Radical Biol. Med., 2008, 45, 549–561.
17. J.-P. R. Chauvin and D. A. Pratt, Angew. Chem., Int. Ed., 2017, 56, 6255–6259.
18. Y. Nishii, M. Ikeda, Y. Hayashi, S. Kawauchi and M. Miura, J. Am. Chem. Soc., 2020, 142, 1621–1629.
19. A. Krȩżel and W. Bal, J. Inorg. Biochem., 2004, 98, 161–166.
20. I. Canals, F. Z. Oumada, M. Rosés and E. Bosch, J. Chromatogr. A, 2001, 911, 191–202.
21. G. Klopman, K. Tsuda, J. B. Louis and R. E. Davis, Tetrahedron, 1970, 26, 4549–4554.
22. C. Storkey, M. J. Davies and D. I. Pattison, Free Radical Biol. Med., 2014, 73, 60–66.
23. X. L. Armesto, M. Canle L, M. I. Fernández, M. V García and J. A. Santaballa, Tetrahedron, 2000, 56, 1103–1109.
24. P. Nagy and M. T. Ashby, J. Am. Chem. Soc., 2007, 129, 14082–14091.
25. N. T. Ioffe, M. I. Kalinkin, I. N. Rozhkov and I. L. Knunyants, Bull. Acad. Sci. USSR, Div. Chem. Sci., 1969, 18, 1929–1930.
26. M. J. Davies and C. L. Hawkins, Free Radical Res., 2000, 33, 719–729.
27. T. P. A. Devasagayam, A. R. Sundquist, P. Di Mascio, S. Kaiser and H. Sies, J. Photochem. Photobiol., B, 1991, 9, 105–116.
28. A. W. M. Nieuwint, J. M. Aubry, F. Arwert, H. Kortbeek, S. Herzberg and H. Joenje, Free Radical Res. Commun., 1985, 1, 1–9.
29. P. Di Mascio and H. Sies, J. Am. Chem. Soc., 1989, 111, 2909–2914.
30. G. R. Buettner, FEBS Lett., 1984, 177, 295–299.
31. A. J. McGrath, G. E. Garrett, L. Valgimigli and D. A. Pratt, J. Am. Chem. Soc., 2010, 132, 16759–16761.
32. T. P. A. Devasagayam, A. R. Sundquist, P. Di Mascio, S. Kaiser and H. Sies, J. Photochem. Photobiol., B, 1991, 9, 105–116.
33. J. A. Montgomery Jr, M. J. Frisch, J. W. Ochterski and G. A. Petersson, J. Chem. Phys., 2000, 112, 6532–6542.
34. C. Chatgilialoglu and M. Guerra, in Sulfur-Centered Reactive Intermediates in Chemistry and Biology, ed. C. Chatgilialoglu and K.-D. Asmus, Springer US, Boston, MA, 1990, pp. 31–36.
35. J. Mönig, K.-D. Asmus, L. G. Forni and R. L. Willson, Int. J. Radiat. Biol. Relat. Stud. Phys., Chem. Med., 1987, 52, 589–602.
36. A. V Kachur, C. J. Koch and J. E. Biaglow, Free Radical Res., 1998, 28, 259–269.
37. H. Berger, Recl. Trav. Chim. Pays-Bas, 1963, 82, 773–789.
38. G. Monteiro, B. B. Horta, D. C. Pimenta, O. Augusto and L. E. S. Netto, Proc. Natl. Acad. Sci. U. S. A., 2007, 104, 4886–4891.
39. L. Turell, H. Botti, S. Carballal, G. Ferrer-Sueta, J. M. Souza, R. Durán, B. A. Freeman, R. Radi and B. Alvarez, Biochemistry, 2008, 47, 358–367.
40. M. Conrad and D. A. Pratt, Nat. Chem. Biol., 2019, 15, 1137–1147.
41. D. D. Saraev, Z. Wu, H.-Y. H. Kim, N. A. Porter and D. A. Pratt, ACS Chem. Biol., 2023, 18, 2073–2081.
42. W. M. Haynes, CRC Handbook of Chemistry and Physics, Taylor & Francis, 93rd edn, 2012.
43. J. N. Cvengroš Stefan, B. Anne and S. Hans-Günther, Synlett, 2008, 2008, 1993–1998.
44. S. Hoops, S. Sahle, R. Gauges, C. Lee, J. Pahle, N. Simus, M. Singhal, L. Xu, P. Mendes and U. Kummer, Bioinformatics, 2006, 22, 3067–3074.
45. J. M. Aubry, J. Am. Chem. Soc., 1985, 107, 5844–5849.
46. P. S. Surdhar and D. A. Armstrong, J. Phys. Chem., 1987, 91, 6532–6537.
47. W. H. Koppenol, D. M. Stanbury and P. L. Bounds, Free Radical Biol. Med., 2010, 49, 317–322.
48. T. P. A. Devasagayam, P. Di Mascio, S. Kaiser and H. Sies, Biochim. Biophys. Acta, Gene Struct. Expression, 1991, 1088, 409–412.
49. V. Anschau, G. Ferrer-Sueta, R. L. Aleixo-Silva, R. Bannitz Fernandes, C. A. Tairum, C. C. C. Tonoli, M. T. Murakami, M. A. de Oliveira and L. E. S. Netto, Free Radical Biol. Med., 2020, 156, 207–216.

---

Footnotes

† Electronic supplementary information (ESI) available: Experimental procedures, characterization data, NMR spectra, additional kinetic data, computational data. See DOI: https://doi.org/10.1039/d4sc04871g

‡ Kinetic studies on reactions of protein sulfenic acids have been conducted, however, the approach did not involve direct visualization of the RSOH intermediate. See ref. 39.

§ It was assumed that the sulfinic acid is fully deprotonated at this pD, based on previously measured pK_a values.¹⁷

¶ Reduction potentials in water (E⁰(RS˙/RS⁻) = +0.84 V vs. NHE⁴⁶ and E⁰(¹O₂/O₂˙⁻) = +0.81 vs. NHE⁴⁷) suggest that the reaction of ¹O₂ and RS⁻ is essentially thermoneutral. This may indicate an even stronger oxidant than ¹O₂ is formed in the reaction. Earlier studies support this notion, as it has been shown that thiols significantly exacerbate the damage to DNA caused by ¹O₂ and that hydroxyl radicals may be formed in the reaction.⁴⁸

|| Three equivalents of ascorbate were selected for this reaction based on the reported rate constants.⁴⁹ If reduction were to occur, it would be possible to monitor by ¹⁹F NMR under these conditions.

---