Accessing Gold π-Acid Reactivity under Electrochemical Anode Oxidation (EAO) through Oxidation Relay

The gold π-acid activation under electrochemical condition is achieved for the first time. While EAO allowing easy access to gold(III) intermediates over alternative chemical oxidation under mild conditions, the reported examples so far limited to coupling reactions due to the rapid AuIII reductive elimination. Using aryl hydrazine-HOTf salt as precursors, the π-activation reaction mode was realized through oxidation relay. Both alkene and alkyne di-functionalization were achieved with excellent functional group compatibility and regioselectivity, which extended the versatility and utility of electrochemical gold redox chemistry for future applications to come.

oxidation 16,17 etc. The recent development of electrochemical anode oxidation (EAO) provided an interesting new strategy in achieving metal oxidation under mild conditions without the requirements of strong chemical oxidants. [18][19][20][21][22][23][24] The challenge was the competing metal cation cathode reduction, which gave rapid catalyst quenching. Through the tuning of electrides and reaction conditions, our group recently reported the rst EAO gold redox catalysis for alkyne oxidative coupling with proton reduction (formation of H 2 ) as the cathode counter reaction 25 (Fig. 1A). Later, this EAO promoted gold redox chemistry was successfully extended to sp-sp 2 coupling under basic condition with O 2 reduction as the cathode counter reaction. 26 While promoting direct C-C coupling using EAO mediated gold redox chemistry is interesting, arguably, the most unique reactivity of cationic gold catalysis over other metal cations is the excellent π-acid activation ability toward C-C multiple bonds (alkene, alkyne and allene). [27][28][29] In both examples mentioned above, no gold-cation π-acid reactivity was observed, even the cationic gold(III) intermediates are clearly formed during the process ( Figure 1B). Therefore, success protocol in accessing gold π-acid reactivity under EAO condition is not only interesting as fundamental catalysis research, but also opens opportunities for new transformations by combining the gold π-activation ability and EAO gold redox chemistry with no interruption from the traditionally required external strong oxidants. In this work, we report the rst example of alkene and alkyne π-acid activation under EAO promoted gold redox catalysis through arylhydrazine assisted oxidation relay ( Figure 1C).
As discussed above, although strong oxidants, Select uor or PIDA, could give gold(I) oxidation, the need of stoichiometric amount of strong oxidative reagents greatly limited the application of this methods for practical synthesis, both for cost and substrate compatibility. Recently, photo or base promoted diazonium salt activation has been developed to generate L-Au III -Ar intermediate upon reacting Au I with the in-situ formed aryl radical. [8][9][10][11][12][13][14][15] While with a lot of successes, this method limited to EWG modi ed arene due to the need of reactive aryl radical for gold(I) oxidation. Bourissou [30][31][32][33] and Patil 34-40 recently reported the ligand enabled gold(I) oxidative addition to aryl iodide with the assistant of silver salts. The hemilabile P-N ligand was applied to facilitate the initial oxidative addition and substrates with electron rich arene could be tolerated under this new condition. While this seminal work showed good reactivity for both cross-coupling and π-activation, the requirement of at least one equivalent of Ag + salts limited the practical application of this method due to low atom economy and concern of functional group tolerability.
With the high e ciency to achieve metal cation oxidation, EAO has been successfully applied into many transition-metal catalysis under controllable cell potential and steady reaction rate. Clearly, the EAO promoted gold redox chemistry is very attractive as it tackled the core challenges in gold redox catalysis, Au I oxidation under mild conditions. The problem, however, is the rapid gold cation reduction on cathode as Au 0 is very easy to form under the redox conditions. Therefore, to facilitate EAO-assisted gold redox catalysis, developing effective cathode counter reactions (over gold cation reduction) is essential. After careful screening of reaction parameters, our group recently discovered that proton hydrogenation could be used as the cathode counter reaction if conducting the reaction under acidic conditions. Based on this mechanistic insight, we reported the alkyne oxidative coupling under EAO condition as shown in Fig. 2A.
With the success on oxidative diyne coupling, we tried to extend this EAO promoted gold redox chemistry into aryl oxidative coupling using ArB(OH) 2 . However, under the similar acidic conditions (MeCN/HOAc solvents), no coupling products were obtained due to rapid L-Au I -Ar protodeauration (Fig. 2B).
Interestingly, after extensive condition screening, we discovered that basic condition could prevent the protodeauration with O 2 reduction (forming HO − ) as the cathode counter reaction. The aryl coupling was successfully achieved under this EAO conditions. It is important to note that these studies not only provided two practical conditions for gold catalyzed C-C coupling under the electrochemical conditions, but also revealed some important mechanistic information of gold redox catalysis under EAO conditions.
Monitoring the reaction with 31 P NMR clearly con rmed that the formation of PPh 3 -Au-C (C = alkynal or aryl) as the gold(I) resting state under both acidic and basic conditions. Interestingly, the reaction of two different aryl boronic acid (para-tBu and para-COOMe) mixtures (1:1) gave dominate homo-coupling of the more electron-rich substrates (from para-tBu substrate). This result clearly suggested that, while the oxidation of L-Au I -Ar is the turnover limiting step, Au I -Ar gave much faster transmetallation over Au III -Ar.
With these mechanistic insights, we put our attention on the much more challenging gold π-acid reactivity under the EAO conditions, which has not been achieved so far in literature. As shown in Fig. 3A, monitoring the reaction of alkene 1a and aryl-boronic acid 2a under the previous optimized EAO conditions (via NMR. 31 P and 19 F) showed the formation of di-aryl coupling product 4aa and phenol 4ab (from arene oxidation), with no alkene conversion and no compound 3a formation observed at all.
The formation of coupling product 4aa con rmed the gold redox cycle under the EAO condition. However, the 100% recovery of alkene 1a suggested no [L-Au III -Ar] π-complex formation (no alkene activation) under this condition. Considering the [Au I -Ar] could give fast transmetallation to the Au III intermediate, which eventually gave the coupling product through rapid reductive elimination, we postulated that to access Au III π-acid reactivity, minimizing the transmetallation on Au III intermediate is crucial. Based on this analysis, an oxidation relay process was proposed as shown in Fig. 3B 41 In that process, the sulfonyl hydrazines were used as the precursors to form sulfonyl radical through oxidative denitrogenation. Encouraged by this process, we postulated that aryl hydrazines might be applied as the aryl electrophile surrogates through the similar electrochemical oxidation. Considering that gold cation is known as carbophilic, either aryl radical or aryl cation could easily convert Au I into Au III -Ar, especially under the electrochemical oxidative conditions. To testify this hypothesis, aryl hydrazine 5a was used to reactive with either terminal alkyne or aryl boronic acid under out previously developed EAO conditions. The corresponding coupling products were successfully observed, con rming the gold redox process under the EAO conditions with this new hydrazine surrogate concept. It is important to note that for both coupling reactions, the homo-coupling Ar 1 -Ar 1 is minor product, which is consistent with our hypothesis that the proposed oxidation relay could bypass the formation of Au I -Ar intermediate (Fig. 3C).
During the preparation of this manuscript, Xie group reported a similar coupling process using aryl hydrazines. 42 Excellent reaction scope along with good functional group tolerability were achieved, highlighting the unique advantage of this EAO promoted gold redox process. However, despite the seminar works of identifying optimal conditions, 10% dppm(AuCl 2 ), 1 eq. Phen, 3 eq. 2,6-di-tBuPy, MeCN, RVC/Ni, again, only coupling products were obtained with no π-acid activation observed even with substrates containing terminal alkyne and intramolecular NuH.
Encouraged by this result, to explore π-acid reactivity based on our design, the reactions between alkene 1a and hydrazine 5a were performed under various EAO-promoted gold redox conditions. As shown in Fig. 4, the initial reaction gave diaryl coupling 4aa as the major product and many un-identi ed byproducts. The desired alkene aminoarylation product 3a was not observed at all, just like the results reported in the recent Xie's work.
Considering that the cathode counter reaction was the reduction of O 2 , the overall reaction was under basic condition due to the accumulated formation of OH -(monitoring the reaction pH con rmed the increasing of basicity over time). We postulate that the coordination between nucleophiles and gold cation under basic conditions could block the coordination site and quench the π-acid activation. To avoid this undesired Nu-Au coordination, we performed the reaction under acidic conditions with the application of hydrazine salts (ArNHNH 2 -HX), with the expectation to convert the cathode counter reaction into proton reduction (formation of H 2 ). Several hydrazine salts were prepared and applied into this EAO condition. Interestingly, while HCl and HBF 4 hydrazine salts gave almost no improvement, reaction with 5a-HOTf gave the aminoarylation product 3a in 16% isolated yield. This result is exciting as it is the rst successful example of gold π-acid activation under EAO conditions. Further tuning the reaction parameters with various solvents, electrodes, electrolytes, and additives revealed the optimal conditions (10% Ph 3 PAuCl, nBu 4 NPF 6 as electrolytes and C/Pt as electrodes), giving the desired aminoarylation product 3a in 92% isolated yields. Results from some representative alternative conditions are summarized in Table 1 (see detailed screening conditions in SI). Table 1, we rst tested different MeCN and MeOH solvent mixture. The 10:1 ratio improved the yield to 32% (entry 1-3). Switching MeOH to other protic solvent led to a lower yield with more HAT product 4ad (which was the major product of hydrazine oxidative decomposition, see entry 12). Aprotic solvents such as DCM cause the fast gold decomposition with Au 0 formation on the cathode with no formation of 3a (entry 5). The PPh 3 P was identi ed as the optimal primary ligand, showing the good stability of PPh 3 AuCl under EAO conditions (monitoring the reaction with 31 P NMR, entry 6). Other tested ligands, such as R 3 P and NHC carbene, gave rapid gold decomposition. Different electrolytes, such as nBu 4 NBF 4 and LiClO 4 , gave minimal in uence on the reactivity (entry 7). Occasionally, when conducting the reaction under higher concentration (2X), slightly improved yield was observed (entry 8). Based on this condition screening, we concluded that the key factor in uencing the yields of this reaction was the concentration of gold catalyst since the reaction turnover limiting step is likely the Au I oxidation to Au III .

As shown in
When putting the gold catalyst at its maximum solubility (0.05 mmol in 5.5 mL solvent), we were pleased to obtain 70% yield of 3a (entry 10). Finally, through the stepwise addition of hydrazines, the desired aminoarylation product 3a was received in 92% isolated yield, with 100% conversion of 1a (entry 11). The control experiments con rmed that both Au catalyst and current were necessary for this transformation (entry [12][13]. It is important to note > 90% of the gold catalyst can be recovered from the reaction mixture by ash column chromatography as Ph 3 PAuCl, which suggested that the oxidation of hydrazine is the initiation of the oxidation relay. With the optimal conditions developed, we explored the reaction scope. The results are summarized in Table 2. As shown in the reaction scope exploration, the monosubstituted alkene substrates worked well for this transformation. Both OH (alcohol and carboxylic acid) and NH (tosyl amine) could serve as effective nucleophile for this transformation, giving the oxyarylation and aminoarylation correspondingly with good yields (3a-3c). For different ring size, both 5-exo-trig and 6-exo-trig cyclization were achieved with good reactivity. No 7-endo-trig cyclization product was observed (see 3d and 3e). The 1,1 or 1,2disubstituted alkenes (Z or E) gave no cyclization with alkene remaining unreacted (see details in SI). This is likely due to the steric hinderance associated with these substrate that reduced the gold(III) coordination and hindered the effective π-activation. Various aryl hydrazines were also tested. To our satisfaction, both EDG and EWG modi ed aryl hydrazines gave the desired oxy-and amino-arylation products in good yields (3f-3o). Notably, the benzylic proton is usually reactive under the EAO condition due to the formation of reactive benzylic radical. To our surprise, the methylphenyl substituted hydrazine works ne under this condition, giving the desired product 3i, though in a lower yield (likely due to the sequential product decomposition). This result suggested hydrazine is the more reactive reductant under this EAO condition and sequential oxidation is plausible with sequential oxidation sequence. The heterocycle substituted 4-pyridinylhydrazine gave no alkene conversion likely due to the rapid decomposition of the in-situ formed pyridinyl radical. Similarly, alkyl hydrazine and Tosyl hydrazine gave no alkene activation product due to the competing side reaction associated with the hydrazine decomposition under the EAO conditions (formation of reactive alkyl radical). The application of ArNHNH 2 -HOTf salts was critical for this transformation as no alkene activation was observed in all these cases if only corresponding ArNHNH 2 was used for the reaction.
It is known that both Au I and Au III are good catalyst toward alkyne (C ≡ C) activation. One potential concern on the Au III promoted alkyne activation is the chemoselectivity between alkyne and resulting alkene. The lack of reactivity with more substituted alkene makes the chemo selective alkyne activation plausible under this EAO conditions. Various alkynes 4 were prepared and applied under this EAO mediated gold redox catalytic conditions (Table 2).
First, terminal alkyne gave cross coupling conjugated diyne product as previously reported. The internal alkynes work well under this EAO condition, giving the corresponding oxo-and amino-arylation products in good yields. The halogen substitution on arene showed little in uence on this transformation, generating the desired products with good yields (5b-5f). This makes the sequential modi cation feasible from well-established aryl halide coupling. The EWG modi ed alkyne gave negative impact on the reaction (5g-5i), likely due to reduced reactivity of the internal alkyne toward gold(III) π-activation. Impressively, although phenol is considered as reactive functional group in anode oxidation, the phenolmodi ed alkyne gave product 5j though with reduced yield. This result, once again, highlighted the unique advantage of electrochemical conditions with the controllable oxidation sequence. Impressively, aliphatic alkyne, despite containing less reactive C ≡ C, could also perform this transformation, giving the desired products (5k, 5l) in good yields. Interestingly, unlike the alkene activation, only 6-endo-dig products (5o, 5p) were obtained as major product. Besides tosyl amine, other protecting groups, such as Ms amine and Cbz carbamide were also suitable nucleophile for this transformation, allowing easier sequential modi cation/protection of the resulting products. Amide (Ac, 5r) protecting group failed to give the deired product due to the rapid substrate decomposition (hydration under acidic condition). Unfortunately, the TIPS modi ed internal alkyne remains unreacted under the optimal condition (5s), which is likely associated with the less feasible nucleophilic addition containing this large sterically hindered modi cation. It is important to note that, under this EAO promoted gold redox condition, no alkyne hydration was observed in all cases, which highlighted the unique reactivity of this reaction protocol with many promising new developments expected in the future. Overall, this electrochemical strategy presents a broad substrate scope for the alkene and alkyne difunctionalization, showing a new strategy for the synthesis of heterocycles with orthogonal reactivity from the typically know protocols.
In summary, we reported here in the rst success examples of the electrochemical promoted gold(I/III) redox chemistry for the alkene/alkyne π-activation. Arylhydrazine was applied as both coupling partner and radical precursor to promote gold oxidation. The acidic condition was critical for this process along with the application of proper anion. The success in achieving the π-acid reactivity of gold redox catalysis under electrochemical conditions not only provided a practical synthetic pathway to achieve the heterocycle derivatives under mild conditions, but also, more importantly, opens new opportunity to facilitate new transformations through this promising new reaction path. Applications of this new EAO mediated gold redox π-acid activation strategy for other challenging transformations are currently undergoing in our lab. Tables   Table 1 and 2      Gold π-activation under EAO conditions.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.