Electro‐Olefination—A Catalyst Free Stereoconvergent Strategy for the Functionalization of Alkenes

Abstract Conventional methods carrying out C(sp2)−C(sp2) bond formations are typically mediated by transition‐metal‐based catalysts. Herein, we conceptualize a complementary avenue to access such bonds by exploiting the potential of electrochemistry in combination with organoboron chemistry. We demonstrate a transition metal catalyst‐free electrocoupling between (hetero)aryls and alkenes through readily available alkenyl‐tri(hetero)aryl borate salts (ATBs) in a stereoconvergent fashion. This unprecedented transformation was investigated theoretically and experimentally and led to a library of functionalized alkenes. The concept was then carried further and applied to the synthesis of the natural product pinosylvin and the derivatization of the steroidal dehydroepiandrosterone (DHEA) scaffold.

Despite its younghistory of only af ew decades, the Suzuki-Miyaura reactioni so ne of the most utilizedr eactionsi n moderno rganic chemistry. [1,2] The palladium-catalyzedc oupling of boronic acids with organohalides was not only awarded with the Nobel prize in 2010, in fact, ar ecent studyr anks the Suzuki-Miyaura coupling as one of the most frequently used reactions (5th place) in medicinalc hemistry. [1] Besides, many other transition-metal-mediated cross-couplings, namely Stille, Heck, Negishi, Sonogashira,H iyama and Kumada are likewise powerful tools to forge new CÀCb onds. [3] Such indispensable strategies undoubtedly display many advantages and have inspired us to challenge the formation of CÀCb onds withoutt he need of the commonly used transition-metal cata-lysts, thus breaking new grounds in the field of cross-coupling reactions. We first started our ambitious concept by replacing the catalyst with an electrochemical setup. Innate advantages, including the use of inexpensive and reusable electrodes, reaction tuneability and scalability do not only rely on the modern and cutting-edge work from Baran,b ut trace back to many othera dvances in electrochemical synthesis sincet he pioneering works of Volta and Faraday in the 19th century. [4] We alreadye mployed electrochemistry to initiatea ryl-aryl bond formation, inspired by the work of Geske [5] and Waldvogel [6] (Scheme1A), introducing new hetero-substituted tetraarylborates alts (TABs).W ed emonstrated that the formation of "unsymmetrical" TABs alts is enabled by at riple ligand exchange reactiono nc ommerciallya vailable organotrifluroborate speciese mploying aryl-Grignard reagents. Submitting those TABs to mild electrochemical oxidation led to the selective formationo fh eterocoupled biaryls (Scheme 1B). [7] As an alternative route to conventional cross coupling reactions,t he catalystf ree Zweifel olefination cannotb en eglected. [8] Thisp owerful methodology enables the stereospecific formationo fa lkenes from the corresponding alkenyl-organoborinates, as exemplified recently by the groups of Aggarwal and Morken (Scheme 1C). [9] In addition, we demonstrated that the logical combination of different organometallic reagents [10] with boron alkoxides could lead to the formation of the required bis-organoborinates in an efficient one-pot process. [11] Based on these findings, we decided to examinet he reactivity of alkenyl-triaryl borate salts (ATBs) to develop an electro-olefination reaction (Scheme 1D).
ATBs (2)a re underexplored salts, the only representative compound being triphenylvinyl boratew hich can be synthesized by treatment of tetravinyltin with triphenylborane. [12] To investigate the electro-olefination ande xpand the structural variety of ATBs, we aimed to simplify their access. Therefore, we built on our previously described strategy for the synthesis of hetero-substituted tetraarylborate salts( TABs), andd ecided to make ATBs accessible by at riple ligand exchange reaction onto the corresponding potassium alkenyl-trifluoro borates 1 (Molander salts), [13] employing ex situ generatedG rignard or organozinc reagents. [14] We anticipated that the removal of an electron through an oxidation process should occur preferentially on the alkenyl moiety,a voiding the energeticallyd isfavoredd earomatization of one of the aryl groups.A saproof of concept, we first synthesized the models ystems 2a [15] and 2b,p ossessing, respectively para-fluorophenyl and phenylm oieties in addition to the b-styryl substituent (Figure1). To describe the change in the electronic structure upon oxidation of 2a and 2b,s pin and charge densities were computed based on Mulliken population analysiso ft he DFT results. Charge densities were additionally computed using the CHargesf rom ELectrostatic Potentials using aG rid-based (ChElPG) method. [14] Blue areas ( Figure 1A) represent positive spin densities after oxidation. Only the alkenyl substituent is selectively oxidized in both cases whereas the chargea nd spin densities of the other aromatic substituents only change insignificantly,c onfirming our assumptions.
The oxidation potentials of ATBs alts 2a and 2b were determined by cyclic voltammetry and compared to the value measured for commercial sodium tetraphenylb orate ( Figure 1B). With af luoridea tom presento ne ach of the aryl groups,a nE ox value of + 0.81 Vv s. SCE was measured for 2a,s imilar to the one of 2c (+ 0.82 Vv s. SCE). However,i nt he absence of electron-withdrawing substituents, the oxidation potential of 2b was decreased to + 0.67 Vv s. SCE. As expected, it can be concludedt hat alkenyl groups are easier to oxidizea nd that the oxidation potentialv aries with the electronic nature of substituentso nt he moieties surrounding the boron atom. From a chemoselectivity perspective, the favorable oxidation of the olefin leaves no other path for the reaction but to transfer one of the remaining aryl moieties, thereby avoiding the undesirable formation of biaryl homocoupling compounds.
2a wasc hosen to test and optimize the reactionc onditions. [14] Inexpensive and reusable glassy carbon electrodes (GCE) proved to deliver the desired stilbene derivative 3a with optimal conversionsi na cetonitrile at 25 8C. Following the transformation by 1 HNMR (Scheme 2) showed that the borate salt 2a is selectively oxidized into product 3a.F ull conversion can be observed after 2.2 Fi n 1 HNMR studies and conversionrate experiments of the electro-olefination using GC revealed that an optimal yield was obtained after 2F .R emarkably, further oxidation resulted in consumption of the reaction product. Althoughn ob iaryl byproduct was detected in 1 HNMR, traces were found in GC. Interestingly,athird minor compound 3ab can be observed, which was identified as the epoxy-stilbene derivative of 3a.T his sider eactionw ill be discussed later with the mechanistic considerations (Scheme 7).
The synthesis of alkenyl-borate salts can be followed by 11 BNMR and proved quantitative when employing either Grignardr eagents or-in cases of sensitivef unctional groups-organozincspecies. [14] Therefore, we started investigating the scope of the transformationu sing borate salts withoutp rior purification. The reaction was first evaluated engaging (E)-alkenyltrifluoroborates 1a-g as startingm aterials in this two-pot sequence. Upon generation of the desired boratei ntermediates, those were treated with an aqueous solution to remove remaining inorganic salts and were subjected to electrochemical oxidation conditions after switching the solventt oa cetonitrile. The results are depicted in Scheme 3. With electron-withdrawing substituentsp resento nt he aryl moieties, (E)-alkenes 3a-b were obtained in up to 69 %y ield over two steps. In the case of p-CN-substituted phenylg roups,t he corresponding organozinc speciesh ad to be employed, loweringt he overall yield of the 2-pot procedure( 3c,2 9%). This consequentd ecrease in yield can be attributedt ot he lower reactivity of organozinc deriva-  (42 and7 4%). Varying the substitution pattern on the alkenyl moietyd id not influencet he course of the reaction, and 3f-g werei solated in 55 to 71 %. Heteroaryl groups were also tolerated in the electro-olefination process, furnishings tructures 3h-j in up to 68 %yield. Interestingly,t risubstituted double bonds also led to the corresponding olefinated aryl derivative 3k in good yield (70 %). The formation of the borates altp roved however difficultw hen an acrylate derivative was used. The introduction of 3-pyridylzinc onto at rifluoroboryl acrylate and subsequente lectro-olefination only gave 25 %o fp roduct 3l.N otably, all derivatives were obtained with excellent (E/Z)r atios, up to 99:1 Z-alkenyl trifluoroborates were employed next. Following the same two-pot protocol, the freshly generated Z-alkenyl-triaryl borates were engaged crude in the electro-olefination under oxidative conditions. Diversely substituted aryl moieties were able to perform the coupling reaction, furnishing compounds 3m-s in reasonable yields (43 to 64 %). It is however interesting to notice that all derivatives were isolateda stransisomers. Given that either of the startingm aterial (E or Z)g ives the same thermodynamic E isomers after electro-coupling, the strategyi ss tereoconvergent (Scheme3). As it will be discussed in the mechanistic part, we assumet hat the oxidation of the double bond into ar adical cationic species allows for the resulting bonding system to freely rotate and adopt the thermo- dynamically more stable configurationb efore abstraction of the boron-containing moiety (Scheme 7).
Next, we applied the method to the derivatization of more challenging structures to demonstrate the synthetic potential of our ATBs alts. Dehydroepiandrosterone( DHEA)w as derivatized into aT BS-protected ether and the carbonyl function transformed into the correspondinga lkenyltrifluoroborate 1o.
The addition of arylmagnesium bromide reagents to 1o,f ollowed by electro-olefination under the optimized oxidative conditions described above furnished functionalized molecules 5a and 5b in up to 70 %y ield (Scheme 5A). In addition, b-styryltrifluoroborate 1a wase mployed as substrate for the synthesis of the natural product pinosylvin (Scheme 5B). 3,5-Dimethoxyphenylmagnesiumb romide was introduced to perform the triple ligand exchange reaction and gave the intermediate alkenyltriaryl borate species. Subsequent electroolefination and demethylation with BBr 3 furnished 5c in 35 % yield over three steps with perfect control of the diastereoselectivity (E/Z = 99:1). Furthermore, the chemoselectivity was investigated on our benchmark salt 2a under distinct oxidative conditions (Scheme 5C). As already mentioned before,t he electro-olefination occurs in as tereoconvergent manner.W e selectively obtain the stilbened erivative 3a using (E)-2a or (Z)-2a in moderate to good yields. In contrast, typical Zweifel conditions led to as tereospecific inversion of the double bond configuration, as the reactionp roceeds through two consecutive stereospecific steps (1,2-metallate rearrangement and antiperiplanar b-elimination). The (Z)-isomer can therefore be synthesized using Zweifelc onditions (Scheme 5C)a nd (Z)-3a was isolated in 86 %y ield (E/Z ratio < 1:99). Noteworthy,s tereodivergent Zweifel protocols have been developed. Even though the presented methodm ight be less versatile than thesec ontributions, our strategy avoids the use of highlytoxic chemicals such as BrCN and PhSeCl. [16] Lastly,w es et out to ascertain the mechanism of this intriguing reaction, building on conversion experiments, cyclovoltammetry and theoretical considerations (Figure 1a nd Scheme2). Crossover experiments were conducted by mixing different borate salts under electrochemical conditions, confirming the absence of products resulting from intermolecular reactions and ruling out the possibility of intermolecular processes. [14] After selectiveo xidationo ft he alkenylm oiety,arearrangement takes place. To study the nature of this rearrangement, we synthesized borate salts containing more than as ingle styryl group (6a-b,S cheme 6), employing styryl-Grignard reagents as (E/Z)-mixtures, and submitted them to our electrocoupling conditions. As ar eference, the desired compound 3a was obtained as the sole compound from 2a.W ith as alt bearing two styryl groups (6a), ap roduct ratio of 73:27 of 3a and the diene 7 was obtained (E/Z = 85:15). This result points out that the transfer of av inyl group is not preferred over the transfer of an aryl group, and therefore indicates that the rearrangementi sm ore likely to go through a s-bond breaking process rather than a p-addition, as for the latter an unfavorable dearomatization has to occur.E xample 6b (possessing three styryl moieties) further supports this hypothesis, as 7 was obtained in 54 %a nd 3a in 46 %G C-ratio. The non-statistical distribution of products 3a and 7 in both experiments also indicatest hat the aryl moiety is-in such cases-a bettert ransferable ligand than the styryl group.
In summary,t he alkenylm oiety is more prone to oxidation than the aryl groups (as concluded from quantum-chemical calculations and selectivity experiments, see Figure 1a nd Scheme 6) and leads to an intermediate alkyl radicalc ationic species[ A]( Scheme 7). We then propose that further intramolecular s-addition of one of the aryl moieties undergoes ar earrangement [17] towards intermediate [B]i nw hich the CÀCa lkyl radicalb ond can freely rotate and lead to the thermodynamically favored trans product (E)-3a.O xygenp robablyi nteracts with the reaction intermediates under formation of structure [C], as 3ab was observed in traces under air andi solated in 37 %y ield when the reaction wasc arried out under oxygen atmosphere. It is however important to note that product 3ab does not come from the oxidation of product 3a under electrochemical conditions, as confirmed by control experiments, indicating ar adicalp athway. [14] Based on cyclovoltametry (Figure 1), galvanostatic experiments( Scheme 2) and our findings in the previous work on biaryl electro-coupling, [7] we assume that no second oxidation has to occur during the formation of the desired product 3a.
In conclusion, we have developed an ew conceptual approach to alkene derivatives through electro-olefination. A simple strategy wasa ssembled for the synthesis of alkenylborate salts (ATBs) through ligand exchanges on potassium trifluoroborates. No purification of these salts was required for the sequence to be pursued and deliver the expected coupling compounds in moderate to good yields under electrochemical oxidation. Such method represents an original ands tereoconvergent alternative to the formation of functionalized olefins, openingn ew ways of thinking about CÀCb ond disconnections.
[ b] Product distribution ratios are determined by GC analysis on crude mixtures withouti solation.H omocoupled biaryls are omitted and not included in the GC-ratios for more clarity.