Rh(iii)-catalyzed synthesis of dibenzo[b,d]pyran-6-ones from aryl ketone O-acetyl oximes and quinones via C-H activation and C-C bond cleavage.

A redox-neutral synthesis of dibenzo[b,d]pyran-6-ones from aryl ketone O-acetyl oximes and quinones has been realized via Rh(iii)-catalyzed cascade C-H activation annulation. A possible Rh(iii)-Rh(v)-Rh(iii) mechanism involving an unprecedented β-C elimination step was proposed. This journal is © The Royal Society of Chemistry.

The dibenzo[b,d]pyran-6-one is one of the most important structural motifs widely present in natural products with pharmacological relevance,[1] such as gut microbiota metabolites urolithins (1–4) that show anti-inflammatory, antiglycative and neuroprotective effects,[2-4] and the extracts of an endophytic fungus Cephalosporium acremonium IFB-E007 (5–7) that have pronounced anticancer activities.[5] In addition, the related heterocyclic structure benzo[d]naphtho[1,2-b]pyran-6-one is found in some bactericidal and antitumor natural products including gilvocarcins[6,7] (8–10) chrysomycins[8,9] (11–13), etc. (Fig. 1). Therefore, a number of approaches to access dibenzo[b,d]pyran-6-ones have been developed via the intra- or inter-molecular biaryl formation as the key step.[10] However, many of these methodologies require multi-step reactions, and the development of new efficient synthetic methods, especially those easy one-step reactions that are still of great interest.

Fig. 1

Selected representative natural products.

In the past decade, transition-metal-catalyzed C–H bond activation has proven to be a powerful tool in organic syntheses[11] and several methods for the synthesis of dibenzo[b,d]pyran-6-ones via C–H activation have been reported.[12] Actually, in 2015, our group reported Rh(iii)-catalyzed synthesis of dibenzo[b,d]pyran-6-ones from N-methoxybenzamides and quinones through C–H activation annulation.[13] Interestingly, we obtained the same products using aryl ketone O-acetyl oximes as substrates to react with quinones under Rh(iii)-catalyzed conditions in this work. Rh(iii)-catalyzed C–H activation using ketoximes as substrates has been developed for synthesis of various substituted heterocycles.[14] Compared to the previous reports, this reaction undergoes a novel mechanism involving an unexpected C–C bond cleavage, which is attractive. Moreover, our study demonstrated that solvent is vital to these reactions. In 2018, we reported Rh(iii)-catalyzed annulation of aryl ketone O-acetyl oximes with quinones to synthesize 6H-benzo[c]chromenes with acetone as a co-solvent.[15] Herein, we described Rh(iii)-catalyzed synthesis of dibenzo[b,d]pyran-6-ones using the same substrates without acetone (Scheme 1).

Scheme 1

Rh(iii)-catalyzed divergent C–H activation annulation with quinones.

Initially, the reaction of acetophenone O-acetyl oxime 1a with benzoquinone 2a was employed to optimize the reaction conditions (Table 1). When the reaction was conducted in the presence of [Cp*RhCl2]2 (2.5 mol%), AgSbF6 (10 mol%) and PivOH (100 mol%) in MeOH at 50 °C for 12 h, 2-hydroxy-6H-dibenzo[b,d]pyran-6-one 3a was obtained in 12% yield (Table 1, entry 1). Elevating the reaction temperature led to a higher yield of 3a (Table 1, entries 1–4). Solvent screening (Table 1, entries 4–9) revealed that reaction in MeOH gave a higher yield of 3a (Table 1, entry 4). Among the additives tested, benzoic acid was the most favorable with respect to product yield (Table 1, entries 4, 10 and 11). Decreasing the amount of benzoic acid to 75 mol% resulted in the best yield of 3a (Table 1, entry 12).

Optimization of the reaction conditionsa

Entry	Additive	Solvent	Temp °C	Yieldb (%)
1	PivOH	MeOH	50	12
2	PivOH	MeOH	70	20
3	PivOH	MeOH	90	36
4	PivOH	MeOH	110	43
5	PivOH	EtOH	110	26
6	PivOH	DMF	110	37
7	PivOH	THF	110	16
8	PivOH	HFIP	110	0
9	PivOH	Acetone	110	Trace
10	HOAc	MeOH	110	Trace
11	Benzoic acid	MeOH	110	50
12c	Benzoic acid	MeOH	110	70
13d	Benzoic acid	MeOH	110	63

Reaction conditions: 1a (0.2 mmol), 2a (0.3 mmol), [Cp*RhCl2]2 (2.5 mol%), additive (100 mol%), solvent (1 mL) for 12 h.

Isolated yields.

Benzoic acid (75 mol%) was added.

Benzoic acid (50 mol%) was added.

Under the obtained optimum reaction conditions above (Table 1, entry 12), we surveyed the reaction scope (Table 2). First, the reactions of various aryl ketone O-acetyl oximes 1 with 2a were examined. For acetophenone O-acetyl oximes, substrates with electron-donating groups or phenyl at the para-position of aryl groups participated well in this reaction and the corresponding products were obtained in good yields (3a–3f). Substrates with halogens or strong electron-withdrawing group trifluoromethyl gave the products in lower yields (3g–3i). Substrate bearing the methyl or chlorine at the meta-position provided the desired products 3l and 3n with exclusive regioselectivity toward the less-hindered site, whereas the meta-methoxy-substituted derivative gave a mixture of regioisomers (3m/3m′ = 2.5 : 1), revealing that the nature of the substituent at the meta-position had an effect on the regioselectivity. 3,4-Disubstituted acetophenone O-acetyl oximes smoothly reacted to result the corresponding dibenzo[b,d]pyran-6-ones 3o and 3p in moderate yields. 2-Acetonaphthone O-acetyl oxime also produced the target product 2-hydroxy-6H-naphtho[2,3-c]chromen-6-one 3q. Next, we examined the reactivity of quinone derivatives with 1a under the established conditions. Methyl benzoquinone afforded the desired molecule in 46% yield with regioisomers (3r/3r′) in a ratio of approximately 1.2 : 1. The naphthoquinone could be also well tolerated, giving benzo[d]naphtho[1,2-b]pyran-6-one 3s in 65% yield. Furthermore, para-methyl-substituted or meta-methyl-substituted acetophenone O-acetyl oximes also smoothly reacted with naphthoquinone to give the corresponding products 3t or 3u. Thus, several tetracyclic benzo[d]naphtho[1,2-b]pyran-6-ones were synthesized successfully.

The reaction scopea

Standard conditions.

To shed light on the reaction mechanism of this annulation, the reaction of acetophenone O-acetyl oxime 1a with benzoquinone 2a under standard conditions was detected by GC-MS, and benzonitrile was observed (detected by GC-MS; see ESI†). This result suggested this reaction might undergo a β-C elimination. Then, deuterium-labeling experiments were further carried out to gain some insights into the catalytic mechanism. A competition between protio and deutero 1a showed a KIE value of 1.86 at early conversion. The KIE was further measured from two side-by-side reactions using protio and deutero 1a with 2a and a KIE value of 2.03 was observed (Scheme 2). These results demonstrated that the C–H bond cleavage process might be involved in the rate-determining step.

Scheme 2

Kinetic isotope effect experiments.

On the basis of our previous work, present observations and literature precedent,[11,13,15,16] a mechanistic pathway is proposed (Scheme 3, taking the reaction of substrate 1a with benzoquinone 2a as an example). First, O-acetyl oxime 1a reacts with the active Cp*Rh(iii) species through directed C–H cleavage to form a five-membered rhodacycle intermediate I. Next, coordination of the benzoquinone affords intermediate II, which undergoes migratory insertion into the incipient Rh–C bond to form a seven-membered rhodacycle III. Protonolysis and aromatization deliver biaryl intermediate IV. Then, an oxidative addition of Rh(iii) into the O–N bond is possible to produce the Rh(v) species V,[17] followed by β-C elimination to give the intermediate VI.[18] A subsequent intramolecular nucleophilic addition of intermediate VI delivers the intermediate VII, which undergoes hydrolysis to generate the final product 3a.

Scheme 3

Proposed mechanism.

Conclusions

In summary, we have developed a novel Rh(iii)-catalyzed cascade C–H activation annulation with readily available and inexpensive substrates for the convenient and direct synthesis of dibenzo[b,d]pyran-6-ones. In this process, we proposed a possible Rh(iii)–Rh(v)–Rh(iii) pathway, which might undergo an unprecedented β-C elimination step. This is the first example of β-C elimination via Rh(iii)-catalyzed C–H bond functionalization. Further studies into the detailed reaction mechanism is ongoing in our laboratory.

Conflicts of interest

There are no conflicts to declare.