Key words
electrochemical amidation - C(sp
3)–H - hexafluoroisopropanol
Amides are prevalent in bioactive molecules. As mentioned by Ertl in the corresponding
review:[1] ‘The most frequent FG in bioactive molecules is the amide, in either its secondary
or tertiary disposition, and this FG is present in 40.3% of all molecules.’ Consequently,
tremendous attention has been devoted to the synthesis of amides. Conventional approaches
commonly require activated acylation reagents.[2] Growing concerns related to sustainable chemistry has shifted attention to the direct
amidation of C(sp3)–H, which would provide an approach to upgrade primary or secondary amides to secondary
or tertiary amides, respectively.
In the area of the intermolecular C(sp3)–H amidation, two main strategies involving C–H activation[3] and nitrene insertion[4] were developed (Scheme [1a]). A range of sophisticated amidating reagents, such as dioxazolones, acyl azides,
and hydroxamate were devised for these transformations. In the sharp contrast, primary
amide as one of the most accessible amidating reagents has received far less attention
due to its weak acidity and nucleophilicity. As depicted in the Scheme [1b], the pK
a
[5] of benzenesulfonamide and benzamide ranges from 16.1 to 23.3.
Scheme 1 Approaches for the C(sp3)–H amidation
The past decade has witnessed the explosive progress in synthetic electrochemistry[6] since it provides distinct and efficient solutions for conventionally challenging
transformations. Electrochemical oxidative amination of C(sp3)–H has received tremendous attention. Sulfonamide as amidating reagent was explored
by some groups. For instance, Lei[7] and coworkers used acidic sulfonimide as an amidating reagent in the reaction with
arenes (Scheme [1c]). A remarkable breakthrough in the site-selective C(sp3)–H amination was recently achieved by the Xu[8] group with benzenesulfonamide as nitrogen source (Scheme [1d]). At almost the same time, Ackermann[9] independently developed an electrochemical approach for allylic C(sp3)–H sulfonamidation (Scheme [1e]). Despite this impressive progress, using benzamide derivatives as a direct amidating
reagent still suffer from formidable challenges. Very recently, we reported a paired-electrolysis
strategy for the electrochemical amidation,[10] in which alkoxyamide was used as the precursor of primary amide (Scheme [1f]). In our line of research[11] in synthetic chemistry, we questioned whether hexafluoroisopropanol (HFIP)[12] could enhance the reactivity of benzamide via proton-coupled electron-transfer (PCET)[13] effect to enable the direct electrochemical C(sp3)–H amidation. Indeed, a direct electrochemical amidation of C(sp3)–H was readily achieved by virtue of unique property[12] of HFIP.
At the outset, we probe the solvent effect of HFIP on the redox property of benzamide
(1a). As shown in the cyclic voltammogram of benzamide (Figure [1a]), a new oxidation peak at 1.12 V was detected upon introducing 1 equivalent of HFIP.
This result suggests that the HFIP could significantly enhance the anodic oxidation
of benzamide as compared to the former peak at 2.09 V. Additionally, a uniform enhancement
effect of HFIP on the oxidation of other amides has also been recorded (see the Supporting
Information for details). To rationalize the effect of HFIP, nuclear magnetic resonance
(1H NMR) study was conducted (Figure [1b]). Mixing HFIP with benzamide lead to obvious variation of peaks of benzamide. Specifically,
peaks of N–H shift to low field (δ = 5.92–6.08 ppm), indicating a hydrogen-bonding
effect between HFIP and benzamide. Taken together, proton-coupled electron transfer
(PCET) was proposed for the enhancement effect of HFIP on the oxidation of benzamide.
Figure 1 Experiment study on the solvent effect of HFIP
Scheme 2 Substrate scope. Reagents and conditions: 1 (0.5 mmol), 2 (0.75 mmol), platinum plate anode (1.5 × 1.5 cm2), graphite rod cathode (0.6 × 10 cm), constant current electrolysis (15 mA, 3 h,
3.3 F/mol), mixed solvent (CH3CN/HFIP = 9/1, v/v), undivided cell. a 5 mmol scale, 75 mA, 6 h.
Having identified the solvent effect of HFIP, we selected xanthene as a reaction partner
with benzamide (Figure [1c]), which would allow a direct access to a broad range of bioactive scaffolds.[14] After a series of optimizations, the electrochemical C(sp3)–H amidation between xanthene and benzamide was readily achieved using mixed solvent
(CH3CN/HFIP), platinum anode, and graphite cathode; the desired product was accessed in
99% yield. Removal of HFIP or replacing it with methanol led to diminished yields.
With the optimal conditions in hand, a broad range of amides was examined to illustrate
the reaction generality (Scheme [2]).[15] First, electronic property effect was explored by using para-substituted benzamides as substrates (3b–h). It showed that the redox-labile iodine group (3f) and the strongly electron-withdrawing cyano group (3h) resulted in lower yields. Second, it was found that changing the para substitution to meta (3i) or ortho (3j) substitution marginally affected the reaction efficiency. Third, other aromatic
amides (3k–m), alkenyl amides (3n–o), and aliphatic amides (3p–q) were also amenable to afford the desired amidation products, although aliphatic
amides led to diminished yields due to its inactive redox property. It is noteworthy
that the bioactive molecule XAA (3o)[14a] can be directly accessed with this electrochemical protocol. Subsequently, secondary
amides (3r–t), carboxamide (3u), phosphinamide (3v), and sulfinamide (3w) were also employed as the amidating reagents, and the corresponding products (3r–w) were delivered in moderate to good yields. To demonstrate the synthetic utility
of this approach, amides derived from natural product (l-proline) and pharmaceuticals (naproxen, probenecid) were subjected to the optimal
conditions. To our delight, the desired amidated xanthenes (3x–z) were successfully accessed in satisfactory yields. Replacing xanthene with thioxanthene
(3aa) caused a significant drop in the reaction yield, while substituted oxanthenes (3ab–ad) were well tolerated with good yields. This result might attribute to the overoxidation
of the product 3aa. Remarkably, this electrochemical protocol can be readily scaled up, and gram-scale
product 3a was readily accessed in 72% yield.
To get insight into the reaction mechanism, control experiments and a cyclic voltammogram
experiment were conducted. As shown in the Figure [2a], some radical scavengers were introduced to the reaction conditions. Obvious suppression
effect was observed, although the desired product 3a still can be accessed in 13–78% yield. This result suggests that the radical species
is involved in this reaction. The kinetic isotope effect was also studied (Figure
[2b]), and it indicates that the cleavage of C(sp3)–H is the rate-determining step in the reaction. Finally, a cyclic voltammogram experiment
(Figure [2c]) showed that closed onset-waves were detected for the substrates benzamide (1a) and xanthene 2a in the mixed solution of acetonitrile and HFIP. This result supports that two substrates
are oxidized simultaneously over the surface of the anode.
Figure 2 Mechanism study
Based on the experimental observations and previous report,[16] a plausible reaction mechanism was proposed (Scheme [3]). Under cathodic reaction, acidic solvent HFIP is reduced to hydrogen and the conjugated
base A. With the PCET assistance of base A, benzamide 1a is oxidized to amidyl radical B. Simultaneously, xanthene 2a proceeds with a sequential single-electron transfer and deprotonation to afford a
persistent radical C, which is immediately intercepted by B to give the final product 3a. Alternatively, radical C can be further oxidized to carbocation D (path II). Under nucleophilic attack of 1a, the desired product 3a is delivered.
Scheme 3 Proposed reaction mechanism
In conclusion, a direct electrochemical amidation between xanthene and benzamides
was reported. In this transformation, a significant enhancement effect of HFIP on
the reaction performance was observed. Proton-coupled electron-transfer effect was
proposed for the role of HFIP in the reaction according to the 1H NMR study. Further investigation on the solvent effect of HFIP is ongoing in our
laboratory.
