Single-Electron-Transfer Oxidation of Trifluoroborates and Silicates with Organic Reagents: A Comparative Study

Abstract

In this report, the single-electron-transfer oxidation of alkyl trifluoroborates and silicates has been studied. Different types of oxidation reagents have been examined, focusing on organic oxidants and particularly the use of dyes in photocatalytic oxidations. Both trifluoroborates and silicates could provide C-centered radicals when using a tritylium salt or the Ledwith–Weitz aminium salt. Photocatalysis with the Fukuzumi reagent suggested that trifluoroborates are more easily oxidized than biscatecholato silicates under these conditions.

Key words

radicals - dyes - photocatalysis - trifluoroborates - silicates - single-electron transfer - oxidation

The single-electron-transfer (SET) oxidation of soft carbanions is a very versatile method to access to C-centered radicals.[1] Among possible candidates, ate complexes based for instance on boron, trifluoroborates being the most popular reagents, have already shown versatile reactivities for the generation of radicals.[2] To a lesser extent, hypervalent biscatecholato silicon species have recently emerged as very promising alternatives to the boron derivatives, avoiding any release of noxious fluorinated byproducts.[3] Their synthesis is known[4] yielding bench-stable compounds,[3] and their high electron density make them suitable candidates for oxidation. In this letter, we provide new elements on the SET oxidation of alkyl trifluoroborates 1 and silicates 3, notably focusing on the use of organic oxidants (Scheme [1]).

Scheme 1Generation of alkyl radicals by SET oxidation of alkyl trifluoroborates and silicates

Our own endeavors in this domain started with the copper(II)-mediated oxidation of alkyl trifluoroborates 1 (Scheme [1]),[5] a previously known reaction[6] but not exploited in radical chemistry. In conditions inspired from the related copper(II) oxidation of alkyl pentafluorosilicates 2 by Kumada and coworkers,[7] a series of alkyl (from primary to tertiary ones) trifluoroborates were engaged in oxidative processes. Postfunctionalization of the resulting radical intermediate was achieved by TEMPO spin trapping, allylation, and conjugate addition.[5]

Following these preliminary reports, we wanted to investigate the use of nonmetallic oxidants. We initially showed that the Dess–Martin periodinane (DMP) could be efficiently used for the oxidation of trifluoroborates.[5] Tritylium tetrafluoroborate, an underexplored oxidant,[8] was also tested with a series of trifluoroborates (Scheme [2]). For reasons which need to be elucidated, DMF did not appear as the best solvent for these oxidations. Gratifyingly, good yields of TEMPO adducts 4 were obtained in Et₂O as solvent with benzyl precursor (4a obtained), but also in secondary (4e and 4g) and primary series (4d,d′).[9] Only tert-butyl precursor 1f failed to give a good yield of product (4f, 25%), presumably for steric reasons. Interestingly, these conditions proved to be compatible with conjugate addition since methyl vinyl ketone (MVK) adduct 5 was isolated in satisfactory yield (63%).

Scheme 2 Stoichiometric oxidation of trifluoroborates by organic reagents. ^a Calculated yield from a mixture with (p-BrC₆H₄)₃N. ^b NMR yield.

We also examined the possibility of using Ledwith–Weitz aminium salt (oxidation potential: 1.06 V vs. SCE)[10] as SET oxidative agent of soft carbanions which, to the best of our knowledge, has never been accomplished. A strong solvent effect (Et₂O vs. DMF) was observed in the oxidation of 1a, respectively 2% vs. 69% of 4a. This led us to pursue our study in DMF with this oxidant. However, even in this solvent, the results proved to be much less satisfying compared to the ones obtained with the tritylium oxidant. Only 27% yield (4e) with the secondary substrate 1e, and no TEMPO adduct in the primary alkyl series.

Next, we investigated the reactivity of biscatecholato pentavalent silicates 3. These substrates are amenable to large-scale synthesis and can be rendered rock stable by complexing the potassium counteranion by the 18-c-6 crown ether.[3a] Benzyl silicate 3a served as a preliminary probe (Scheme [3]). It was submitted in Et₂O and DMF to one equivalent of tritylium and aminium. In both solvents, tritylium gave poor yields of 4a (< 20%). However, the use of the aminium salt was more rewarding (86% of 4a in DMF, 16% in Et₂O).[11] This oxidant proved to be competent in DMF for secondary and primary alkyl substrates giving, respectively, 44% of 4h and 61% of 4d,d′. Tritylium can also be used as a reliable alternative oxidant for the silicates 3.

Scheme 3Stoichiometric oxidation of biscatecholato silicates by organic­ reagents. ^a Calculated yield from a mixture with (p-BrC₆H₄)₃N. ^b NMR yield.

Because of its mild conditions and high substrate tolerance, visible-light photocatalytic oxidation was the obvious next step.[12] In the case of trifluoroborates, several groups have established the feasibility of this transformation using ruthenium(II)- or iridium(III)-based photocatalysts.[13] Of note, the resulting radicals can be engaged in photoredox/nickel dual catalysis.[13g] [h] [i] [j] [k] While pentafluorosilicates 2 failed in our hands to undergo any oxidation,[14] we recently showed that biscatecholato silicates constitute advantageous alternatives to the trifluoroborates since they allow upon iridium(III) {Ir[(dF(CF₃)ppy)₂(bpy)](PF₆)} photocatalysis the generation of very unstabilized primary radicals, also successfully engaged in photoredox/nickel dual catalysis.[3]

Scheme 4Photocatalytic oxidation of trifluoroborates and biscatecholato silicates by organic dyes.

Herein, we wanted to examine the possibility to use organic dyes[12] [15] as possible catalysts for the oxidation of these soft carbanions. Based on their frequent use, the following dyes were considered: eosin Y, fluorescein,[16] and Fukuzumi acridinium as catalysts.[17] A preliminary screening with benzyltrifluoroborate 1a showed that the Fukuzumi catalyst was by far the best one (Scheme [4]).

Similar behavior was observed for 3a. Therefore we kept this catalyst for further testing. Both substrate families showed the same trend, that is, the less stabilized is the generated radical, the lower is the yield. Thus, for trifluoroborates, a gradual decrease of yield was observed from benzyl product 4a to least stabilized primary radical adducts 4d,4d′. One could argue that 1g, a secondary substrate, should have given a better yield. But in that case, the final radical is a tertiary one which may undergo competitive pathways and lead to only 18% of 4g. In the case of silicates 3, only stabilized benzyl and allyl radicals could be generated (66% for 4a, 31% for 4b). Interestingly, allyltrifluoroborate 1b and allylsilicate 3b provided 4b in close yields (38% vs. 31%). In sharp contrast, however, secondary trifluoroborates could give TEMPO adducts 4e and 4i contrary to secondary silicate 3h (no 4h formed).[18]

A direct correlation of these findings with redox potentials is not obvious. Oxidation potentials for trifluoroborates span from 1.1 V (benzyl et alkoxymethyl) to 1.83 V vs. SCE (primary and aryl)[13a] [b] [12s] while they have been determined to range from 0.61 V for benzylsilicate 3a to 0.75 V vs. SCE for 3d.[3] Some other key factors are at play in these reactions that we will try to uncover. In all the successful oxidations, TEMPO would act as a sacrificial oxidant to regenerate the photocatalyst and sustain the photocatalytic cycle in agreement with the literature data.[3a] [13a] [19]

In conclusion, this study shows the unprecedented oxidation of trifluoroborates and silicates with a tritylium or an aminium salt as stoichiometric oxidant to generate C-centered radicals. Photocatalytic oxidation could also be achieved with the Fukuzumi acridinium showing a higher reactivity of trifluoroborates than silicates in these conditions. Studies are ongoing to improve silicates photooxidation with organic dyes. The effect of the silyl substituents will notably be studied.

Acknowledgment

We warmly thank CNRS, UPMC, UHA, IUF, MSER (ASN PhD grant to CL), ANR CREDOX, LABEX MiChem (ANR-11-IDEX-0004-02), La Région Martinique (PhD grant to LC), ANR NHCX (11-BS07-008, postdoc grant to VC). COST Action CM1201 is gratefully acknowledged. We thank Professor Kirsten Zeitler (U. Leipzig) for helpful discussions.

• References and Notes

• 1a Dalko PI. Tetrahedron 1995; 51: 7579
• 1b Jahn U. Radicals in Synthesis III. In Topics in Current Chemistry. Vol. 320. Heinrich M, Gansäuer A. Wiley-VCH; Weinheim: 2012: 121
• 1c Jahn U. Radicals in Synthesis III. In Topics in Current Chemistry. Vol. 320. Heinrich M, Gansäuer A. Wiley-VCH; Weinheim: 2012: 191
• 1d Jahn U. Radicals in Synthesis III. In Topics in Current Chemistry. Vol. 320. Heinrich M, Gansäuer A. Wiley-VCH; Weinheim: 2012: 323
• 1e Gansäuer A, Bluhm H. Chem. Rev. 2000; 100: 2771
• 2a Schuster GB. Pure Appl. Chem. 1990; 62: 1565
• 2b Shundrin LA, Bardin VV, Frohn H.-J. Z. Anorg. Allg. Chem. 2004; 630: 1253
• 2c Molander GA, Colombel V, Braz VA. Org. Lett. 2011; 13: 1852
• 2d Lockner JW, Dixon DD, Risgaard R, Baran PS. Org. Lett. 2011; 13: 5628
• 2e Fujiwara Y, Domingo V, Seiple IB, Gianatassio R, Bel MD, Baran PS. J. Am. Chem. Soc. 2011; 133: 3292
• 2f Liwosz TW, Chemler SR. Org. Lett. 2013; 15: 3034
• 2g Neufeldt SR, Seigerman CK, Sanford MS. Org. Lett. 2013; 15: 2302

For the oxidation of boronic acids, see
• 2h Brown HC, Hébert NC, Snyder CH. J. Am. Chem. Soc. 1961; 83: 1001
• 2i Demir AS, Reis Ö, Emrullahoglu M. J. Org. Chem. 2003; 68: 578
• 2j Dickschat A, Studer A. Org. Lett. 2010; 12: 3972
• 2k Tobisu M, Koh K, Furukawa T, Chatani N. Angew. Chem. Int. Ed. 2012; 51: 11363
• 3a Corcé V, Chamoreau LM, Derat E, Goddard J.-P, Ollivier C, Fensterbank L. Angew. Chem. Int. Ed. 2015; 54: 11414

During the course of our investigation, the following complementary report appeared, see:
• 3b Jouffroy M, Primer DN, Molander GA. J. Am. Chem. Soc. 2016; 138 in press; DOI: 10.1021/jacs.5b10963
• 4a Holmes RR. Chem. Rev. 1990; 90: 17
• 4b Chuit C, Corriu RJ. P, Reye C, Young JC. Chem. Rev. 1993; 93: 1371
• 5 Sorin G, Mallorquin RM, Contie Y, Baralle A, Malacria M, Goddard J-P, Fensterbank L. Angew. Chem. Int. Ed. 2010; 49: 8721
• 6a Nishigaichi Y, Orimi T, Takuwa A. J. Organomet. Chem. 2009; 694: 3837
• 6b Carzola C, Metay E, Andrioletti B, Lemaire M. Tetrahedron Lett. 2009; 50: 6855
• 7 For seminal work, see: Yoshida J.-I, Tamao K, Kakui T, Kurita A, Murata M, Yamada K, Kumada M. Organometallics 1982; 1: 369

Tritylium is known as a hydride abstractor, for a recent application, see:
• 8a Xie Z, Liu L, Chen W, Zheng H, Xu QH, Yuan H, Lou H. Angew. Chem. Int. Ed. 2014; 53: 3904 ; and references cited therein

It has been used as a sacrificial electron acceptor in photoredox catalysis, see:
• 8b Daniel M, Fensterbank L, Goddard J.-P, Ollivier C. Org. Chem. Front. 2014; 1: 551
• 9 To a Schlenk flask was added potassium 5-hexenyl-1-trifluoroborate (1d) or potassium [18-crown-6] bis(catecholato)-5-hexenyl-1-silicate (3d, 0.3 mmol, 1 equiv), the oxidizing agent (0.3 mmol, 1 equiv), and TEMPO (0.9 mmol, 141 mg, 3 equiv). The Schlenk flask was sealed with a rubber septum, and evacuated–purged with vacuum–argon three times. Degassed Et₂O or DMF (3 mL) was introduced followed by two freeze–pump–thaw cycles. The reaction mixture was stirred at room temperature for 24 h under an argon atmosphere. The reaction mixture was diluted with Et₂O (50 mL), washed with H₂O or NaHCO_3- (2×), brine (2×), dried over MgSO₄, and evaporated under reduced pressure. The reaction residue was purified by flash column chromatography on silica gel to afford an inseparable mixture of 4d and 4d′ in a 9:1 to 10:1 ratio and an overall yield (37–61%) depending on the oxidizing agent. Compound 4d: ¹H NMR (400 MHz, CDCl₃): δ = 5.82 (m, 1 H), 5.01 (m, 1 H), 4.94 (m, 1 H), 3.73 (t, J = 6.1 Hz, 2 H), 2.07 (q, J = 7.2 Hz, 2 H), 1.55–1.20 (m, 10 H), 1.14 (s, 6 H), 1.09 (s, 6 H) ppm. ¹³C NMR (100 MHz, CDCl₃): δ = 139.1, 114.5, 80.8, 59.9, 59.8, 39.7, 34.0, 33.2, 28.4, 25.9, 20.3, 17.3 ppm. Compound 4d′: ¹H NMR (400 MHz, CDCl₃): characteristic signal at δ = 3.64 ppm (CH ₂O). ¹³C NMR (100 MHz, CDCl₃): characteristic signal at δ = 76.7 ppm (CH₂O).
• 10a Herath AC, Becker JY. J. Electroanal. Chem. 2008; 619-620: 98
• 10b Brinkhaus KH. G, Steckhan E, Schmidt W. Acta Chem. Scand., Ser. B 1983; 37: 499
• 10c Wend R, Steckhan E. Electrochim. Acta 1983; 42: 2027

For a recent use, see:
• 10d Drew SL, Lawrence AL, Sherburn MS. Angew. Chem. Int. Ed. 2013; 52: 4221

(e) For a review, see: Jia, X. Synthesis 2016, 48, 18.
• 11 In comparison, oxidation with 1 equiv of Cu(OAc)₂ gave 45% yield of 4a and with 1 equiv of DMP, 26% of 4a.

For selected reviews on visible-light photoredox catalysis, see:
• 12a Zeitler K. Angew. Chem. Int. Ed. 2009; 48: 9785
• 12b Yoon TP, Ischay MA, Du J. Nat. Chem. 2010; 2: 527
• 12c Teplý F. Collect. Czech. Chem. Commun. 2011; 76: 859
• 12d Narayanaman JM. R, Stephenson CR. J. Chem. Soc. Rev. 2011; 40: 102
• 12e Tucker JW, Stephenson CR. J. J. Org. Chem. 2012; 77: 1617
• 12f Xuan J, Xiao W.-J. Angew. Chem. Int. Ed. 2012; 51: 6828
• 12g Ischay MA, Yoon TP. Eur. J. Org. Chem. 2012; 3359
• 12h Maity S, Zheng N. Synlett 2012; 23: 1851
• 12i Shi L, Xia W. Chem. Soc. Rev. 2012; 41: 7687
• 12j Xi Y, Yi H, Lei A. Org. Biomol. Chem. 2013; 11: 2387
• 12k Prier CK, Rankic DA, MacMillan DW. C. Chem. Rev. 2013; 113: 5322
• 12l Hari DP, König B. Angew. Chem. Int. Ed. 2013; 52: 4734
• 12m Reckenthäler M, Griesbeck AG. Adv. Synth. Catal. 2013; 355: 2727
• 12n Koike T, Akita M. Synlett 2013; 24: 2492
• 12o Xuan J, Lu L.-Q, Chen J.-R, Xiao W.-J. Eur. J. Org. Chem. 2013; 6755
• 12p Zou Y.-Q, Chen J.-R, Xiao W.-J. Angew. Chem. Int. Ed. 2013; 52: 11701
• 12q Hu J, Wang J, Nguyen TH, Zheng N. Beilstein J. Org. Chem. 2013; 9: 1977
• 12r Xie J, Jin H, Xu P, Zhu C. Tetrahedron Lett. 2014; 55: 36
• 12s Koike T, Akita M. Inorg. Chem. Front. 2014; 1: 562
• 12t Hopkinson MN, Sahoo B, Li J.-L, Glorius F. Chem. Eur. J. 2014; 20: 3874
• 12u Schultz DM, Yoon TP. Science 2014; 343: 985

For recent books, see:
• 12v Chemical Photocatalysis. König B. DeGruyter; Berlin: 2013
• 12w Photochemically Generated Intermediates in Synthesis. Albini A, Fagnoni M. John Wiley and Sons; Hoboken: 2013
• 13a Yasu Y, Koike T, Akita M. Adv. Synth. Catal. 2012; 354: 3414
• 13b Miyazawa K, Yasu Y, Koike T, Akita M. Chem. Commun. 2013; 49: 7249
• 13c Koike T, Akita M. Synlett 2013; 24: 2492
• 13d Miyazawa K, Koike T, Akita M. Adv. Synth. Catal. 2014; 356: 2749
• 13e Li Y, Miyazawa K, Koike T, Akita M. Org. Chem. Front. 2015; 2: 319
• 13f Huang H, Zhang G, Gong L, Zhang S, Chen Y. J. Am. Chem. Soc. 2014; 136: 2280
• 13g Tellis JC, Primer DN, Molander GA. Science 2014; 345: 433
• 13h Primer DN, Karakaya I, Tellis JC, Molander GA. J. Am. Chem. Soc. 2015; 137: 2195
• 13i Gutierrez O, Tellis JC, Primer DN, Molander GA, Kozlowski MC. J. Am. Chem. Soc. 2015; 137: 4896
• 13j Karakaya I, Primer DN, Molander GA. Org. Lett. 2015; 17: 3294
• 13k Yamashita Y, Tellis JC, Molander GA. Proc. Natl. Acad. Sci. U.S.A. 2015; 112: 12026
• 13l Huang H, Jia K, Chen Y. Angew. Chem. Int. Ed. 2015; 54: 1881
• 14 Allyl-, cyclopentyl-, t-BuSiF₅K₂ did not give any TEMPO adduct 4 in the following conditions {2 mol% Ir[(dF(CF₃)ppy)₂(bpy)](PF₆), acetone or DMF, TEMPO (2.5 equiv), blue LED}.
• 15a Fukuzumi S, Ohkubo K. Org. Biomol. Chem. 2014; 12: 6059
• 15b Nicewicz DA, Nguyen TM. ACS Catal. 2014; 4: 355
• 15c Ravelli D, Fagnoni M. ChemCatChem 2012; 4: 169

For a recent use, see:
• 15d Griffin JD, Zeller MA, Nicewicz DA. J. Am. Chem. Soc. 2015; 137: 11340
• 16 Zhang X.-F, Zhang I, Liu L. Photochem. Photobiol. 2010; 86: 492
• 17a Fukuzumi S, Kotani H, Okhubo K, Ogo S, Tkachenko NV, Lemmetyinen H. J. Am. Chem. Soc. 2004; 126: 1600
• 17b Benniston AC, Harriman A, Li P, Rostron JP, van Ramesdonk HJ, Groeneveld MM, Zhang H, Verhoeven JW. J. Am. Chem. Soc. 2005; 127: 16054
• 18 To a Schlenk flask were added the organotrifluoroborate 1 or organosilicate 3 (0.3 mmol, 1 equiv), 9-mesityl-10-methylacridinium perchlorate as photocatalyst (0.03 mmol, 10 mol%), and TEMPO (0.66 mmol, 2.2 equiv.). The Schlenk flask was sealed with a rubber septum and evacuated–purged with vacuum–argon three times. Degassed DMF (3 mL) was introduced followed by two freeze–pump–thaw cycles. The reaction mixture was stirred under blue LEDs irradiation at room temperature for 24 h under an argon atmosphere. The reaction mixture was diluted with Et₂O (50 mL), washed with sat. NaHCO₃ (2×), brine (2×), dried over MgSO₄, and evaporated under reduced pressure. The reaction residue was purified by flash column chromatography on silica gel.
• 19 The generated TEMPO N-oxide anion could be silylated or borylated. The resulting anionic products would be eliminated during the aqueous workup. We thank one of the referees for this suggestion.