Vanadium-catalyzed Oxidative Intramolecular Coupling of Tethered Phenols: Formation of Phenol-dienone Products

Abstract

A mild and efficient method for the vanadium-catalyzed, intramolecular coupling of tethered free phenols is described. Corresponding phenol-dienone products are prepared directly in good yields with low loadings of catalyst. Electronically diverse tethered phenol precursors are well tolerated, and the catalytic method was effectively applied as the key step in syntheses of three natural products and a synthetically useful morphinan alkaloid precursor.

Graphical Abstract

Bisphenols constitute an important class of molecules with utility as both complex molecule precursors and as biologically active natural products.^1,2 Such biologically active compounds often incorporate the bisphenol biaryl motif as part of a polycyclic system (Figure 1A).³ Other phenol-containing natural products arise from the direct oxidative coupling of tethered free phenols to afford a phenol-dienone motif (Figure 1B,C).^4,5 Since the discovery that direct phenol oxidation plays a critical role in the biosynthesis of such compounds,^6–8 analogous reagent-based approaches have been of considerable interest for their potential application in the synthesis of plant metabolites.

Figure 1.

Compounds derived from linked phenols.

There are several reports on intermolecular oxidative phenol coupling including transition metal-mediated^9–13 and electrochemical^14,15 phenol homo- coupling and cross-coupling reactions. Despite this, the intramolecular coupling of tethered free phenols for the formation of phenol-dienone products is not well studied. Schwartz et al. disclosed one of the first examples of this transformation, coupling propyl-tethered phenols with stoichiometric VOCl₃ in good yield (Scheme 1a).¹⁶ This work with stoichiometric VOCl₃ was also utilized in the formation of N-(ethoxycarbonyl)-2-hydroxynorsalutaridine,¹⁷ an analog of salutaridine (B, Figure 1), a key intermediate in the biosynthesis of morphine. In addition to V(V), phenol-dienone products have been prepared directly from tethered free phenols using stoichiometric Fe(III)^18,19 or Tl(III)^20,21 as the oxidant. Despite success with these stoichiometric reagents, limitations include air-sensitivity of VOCl₃, toxicity of Tl(III), the need for a significant excess of oxidant, and reduced yields in converting electron-neutral and electron-deficient substrates. A catalytic variant of this direct transformation using stable, easily accessible reagents would therefore be of great utility from the perspective of atom economy and accessing structurally diverse phenol-dienone substrates.

Scheme 1.

Intramolecular Phenol-Type Couplings

An alternative approach for the formation of phenol-dienone products proceeds via oxidation of the corresponding tethered aryl ethers or mixed phenol/ether species, followed by subsequent deprotection. Considerable work in this area has shown that stoichiometric Mo(V)²² and I(III)²³ are competent in the intramolecular oxidative coupling of tethered aryl ethers. The stoichiometric nature of the I(III)-mediated transformation was addressed by Kita et al. in 2008, as they coupled mixed tethered phenol/ethers with catalytic amounts of aryl iodide, which is activated to I(III) by H₂O₂ in the presence of trifluoroacetic acid (Scheme 1b).²⁴ A complementary approach was also disclosed by Wang et al., wherein they oxidized similar substrates using a catalytic system of sodium nitrite in the presence of air and Brønsted acid (Scheme 1c).²⁵ While both of these reports reflect catalytic variants of the intramolecular coupling of tethered arenes, both require the use of electron rich substrates. Additionally, both approaches utilize mixed tethered phenol/ether substrates, which precludes the direct formation of the phenol-dienone motif. Deprotection of aryl ethers to generate the phenol-dienones can be challenging in the presence of the newly formed dienone motif, which is sensitive to acidic and reducing conditions. Orthogonality of protecting groups also becomes a challenge when multiple ether moieties are present in the ether-dienone product.

An important distinction should be made between bis-ether, mixed ether/phenol, and bisphenol intramolecular coupling reactions. These transformations are fundamentally different and are proposed to proceed via different mechanisms. In the case of ether systems, a polar, two-electron mechanism invoking a cationic intermediate is often proposed. In the case of tethered bisphenol cyclizations to form phenol-dienone products, the reaction is more likely to proceed through sequential single electron oxidation of both coupling partners, akin to the mechanisms proposed in the biosynthesis of phenol-derived polycyclic natural products. To our knowledge, there have been no reports detailing the catalytic oxidation of tethered free phenols for the formation of phenol-dienone products.

Inspired by the work of Schwartz^16,17 and our prior work with vanadium Schiff base catalysts,^12,26 we envisioned the development of a catalytic system for the oxidation of tethered free phenols to form phenol-dienone products directly (Scheme 1d). Initial reaction studies were conducted with substrate 1a, 20 mol % vanadium catalyst V1, and O₂ as the co-oxidant (Table 1). When 1,2-dichloroethane (DCE) was utilized as solvent, there was little consumption of 1a and the catalyst system failed to form 2a (Table 1, entry 1). 1,1,1,3,3,3-Hexafluoroisopropanol (HFIP) was identified as a suitable solvent for this reaction because of its unique ability to stabilize electron-deficient reaction intermediates.²⁷ The reaction in HFIP showed good conversion to 2a over 24 h (Table 1, entry 2), with full conversion observed in 48 h affording an isolated yield of 89% (Table 1, entry 3). Interestingly, only para-para coupled products were observed in this reaction, despite the fact that prior work with this catalyst uncovered a preference for ortho-coupling.¹² We hypothesize that the ortho-para coupled product is not formed because of unfavorable steric interactions between phenoxy-bound vanadium and the other aryl partner during the C-C bond-forming step.

Table 1.

Optimization of Reaction Conditions^a

entry	X	loading (mol %)	time (h)	2a (%)b
1c	OEt	20	24	< 5
2	OEt	20	24	56
3	OEt	20	48	(89)d
4	OEt	10	18	24
5	Ot-Bu	10	18	12
6	F	10	18	< 5
7	OCH(CF3)2	10	18	70
8	OCH(CF3)2	10	24	(80)d
9	OCH(CF3)2	5	24	31

^aReaction conditions: 1a (0.10 mmol), V (10–20 mol %), HFIP (0.1 M), O₂ (1 atm) for 18–48 h.

^bDetermined by ¹H NMR spectroscopy with an internal standard (4,4’-di-tert-butylbiphenyl).

^cSolvent = DCE.

^dIsolated yield.

Efforts to reduce reaction time and catalyst loading focused on the catalyst counterion. Vanadium Schiff base catalysts bearing the tert-butoxy and fluoride counterions were shown to form product more slowly over 18 h than the corresponding ethoxy catalyst over the same time period (Table 1, entry 5 & 6 vs entry 4). The observation that catalyst activity tracks with lability of the counterion led to the preparation of a catalyst bearing HFIP as the counterion. Gratifyingly, V2 proved to be more effective in the transformation, forming 2a in 70% yield over the same reaction time (Table 1, entry 7). Reaction for a full 24 hours with 10 mol % V2 led to full conversion of 1a and an 80% isolated yield of 2a (Table 1, entry 8). A reaction profile for the conversion of 1a to 2a with V2 (see Supporting Information, control A) is consistent with first order decomposition through 90% conversion with neither a significant burst nor a significant induction period. Independent experiments where 2a was added in variable amounts prior to reaction of 1a indicate that no product inhibition is occurring, since 1a is converted to the same extent with various excesses of 2a. Even so, further reduction of catalyst loading led to poor conversion over the 24 h reaction period (Table 1, entry 9). A screen of solvents (SI, Table S2) and standard oxidants (SI, Table S3) revealed that HFIP was the optimal solvent for the transformation and that O₂ is the most effective oxidant, outperforming an atmosphere of air, inorganic persulfates, and organic peroxides.

With the optimized conditions for the formation of phenol-dienone products in hand, the substrate scope of this transformation was examined (Scheme 2). In the simplest, unactivated case, the unsubstituted precursor was oxidized in good yield to its corresponding cyclized product (2a). In the case of substitution on the ring where the tether is attached at C3 (red ring in Scheme 2), electron-releasing and electron-withdrawing groups were well-tolerated (2b-2h). Bromo- (1d) and chloro-containing (1e) substrates were oxidized to phenol-dienone product in fair yield requiring an increased catalyst loading. Additionally, substitution is tolerated at all positions of this ring with no loss in coupling efficiency (2c, 2g, 2h). A direct assessment against prior art was undertaken to further validate the method: 1) with Tl(O₂CCF₃),²¹ a 50% yield of 2c was obtained with 79% consumption of 1c; 2) with VOCl₃,¹⁶ a 39% yield was obtained after 85% consumption and several other byproducts were observed; 3) with the method from this manuscript, 76% yield was obtained with 85% consumption. Overall, our catalytic method was easier to undertake (Tl and VOCl₃ need special handling), gave higher yield, provided a much cleaner reaction profile, and resulted in more facile purification.

Scheme 2. Scope of V-catalyzed Intramolecular Coupling of Tethered Free Phenols ^a.

^a Reaction conditions: 1 (0.10–0.20 mmol), V2 (10 mol %), HFIP (0.1 M), O₂ (1 atm) for 24–44 h; yields of isolated material. ^b On a larger scale (5.00 mmol, 1.14 g). ^c 20 mol % catalyst used. ^d Solvent = HFIP/DCE (0.1 M, 1:1). ^e Yield in parentheses based on recovery of starting material.

In the case of substitution on the ring where the tether is attached at C4’ (blue ring in Scheme 2), similar results were obtained (2i-2p). Electron-releasing and electron-withdrawing groups were well tolerated. Notably, dimethoxy substituted 1p was converted to its corresponding highly substituted phenol-dienone product in good yield. Fluoro-substituted 2n was also obtained in fair yield, revealing that deactivated substrates are capable of undergoing oxidation by V2 to form product. Additionally, substitution is tolerated at other positions of this ring (2m, 2o) with no loss in efficiency. Even though V2 is a chiral catalyst, no enantioenduction was observed for 2j-2m, likely due to the catalyst binding phenols that are distal to the site of C-C bond formation. The overall substrate scope establishes that this catalyst system is capable of oxidizing a wide variety of tethered phenol compounds and that the transformation is not limited by a need for electron-rich, nucleophilic arenes as is the case for other methods.^24,25 Efforts to increase reaction scale were successful, with a gram-scale reaction of 2a giving a nearly identical outcome.

In order to further outline the utility of this method, several natural products were identified as synthetic targets due to the presence of the phenol-dienone moiety in their core structure.^17,21,28,29 The natural product pulchelstyrene D (3)²⁸ differs from the simple substrates prepared above (Scheme 2) only in the unsaturation of the propyl tether. Attempts at preparing a version of the substrate with an alkenyl tether were unsuccessful, so an alternative route through the saturated precursor was envisioned. As a result, dimethoxy-substituted substrate 1q was coupled in good yield to generate saturated phenol-dienone product 2q (Scheme 3a). Efforts to directly convert 2q into the natural product 3 were unsuccessful. To overcome this issue, 2q was subjected to a three-step sequence of acetylation, radical-mediated bromination/elimination, and deacetylation to form natural product 3 in 34% yield over three steps (see SI). This synthetic effort represents the first preparation of this naturally occurring compound.

Scheme 3.

Formation of Spiro-dienone Natural Products via Intramolecular Oxidative Coupling: Pulchelstyrene D and Spirolouveline

A further natural product, spirolouveline (4),²⁹ is also closely related to the simple substrates prepared in Scheme 2. In addition to a single methoxy group, the compound contains a free hydroxyl group on the dienone moiety. To directly access 4, a catechol-containing precursor was prepared and subjected to the vanadium-catalyzed oxidation conditions. Unfortunately, a poor reaction profile was observed, likely due to the presence of multiple oxidizable phenolic hydroxyl groups. An alternative approach utilizing a protected form of the catechol was envisioned, and 1r was coupled successfully in good yield to generate bis-alkoxy phenol-dienone 2r (Scheme 3b). Selective removal of the more labile iso-propoxy group of 2r with Lewis acid furnished racemic 4. This synthetic effort also represents the first reported preparation of this naturally occurring compound.

The dracaenone natural products represent a tetracyclic variant of the tethered phenol-dienone motif. The natural product 10-hydroxy-11-methoxydracaenone (6) has been prepared previously via coupling of a free phenol^20,21 with an aryl ether and then by deprotection of the coupled aryl ether²¹ in 37% and 62% (over 2 steps), respectively. In each case, stoichiometric oxidant is required for the key intramolecular coupling step. Using the optimized conditions with V2, natural product 6 was obtained in a single step from its corresponding free phenol precursor in 78% yield (Scheme 4a). This preparation represents a protecting group-free synthesis of 6 with improved efficiency in the key intramolecular phenol coupling step.

Scheme 4.

Formation of Bicyclic Systems via Intramolecular Oxidative Coupling: Dracaenone Natural Product and Salutaridine Analog

Lastly, a synthetically useful salutaridine derivative was prepared using the newly developed catalytic method (Scheme 4b). Reticuline derivative 7 was coupled directly to form salutaridine phenol-dienone product 8 in good yield. Remarkably, only small quantities of the ortho-ortho coupled aporphine-type product were observed. It has been hypothesized that products of this type are not observed due to an unfavorable peri-interaction between the approaching hydroxyl groups in the C-C bond-forming step.¹⁷ Compound 8 is of significant interest in the context of alkaloid synthesis,³⁰ since the coupling reaction establishes the B ring of the classic morphine scaffold with limited pre-functionalization. Considerable precedent exists for conversion of these phenol-dienone products to downstream morphinan alkaloids.^17,31,32 The efficiency of the coupling reaction to form 8 compares well to the stoichiometric variant reported by Schwartz et al.¹⁷ and offers a complementary approach to state-of-the-art electrochemical couplings of the related aryl ethers.^31,32

A series of control experiments were conducted to aid in understanding the mechanism of the vanadium-catalyzed coupling reaction (see Supporting Information). Reactions run without V2 in the presence of dioxygen showed limited conversion of substrate and no product formation, indicating that catalyst is key in oxidizing the substrate (SI, Control C). Additionally, a study concerning catalyst ligand identity (SI, Control E) showed that electron-rich ligands (V1, with t-Bu in place of NO₂) led to slower reaction rates. This result points away from catalyst re-oxidation [V(IV) to V(V)] as rate-limiting and suggests substrate-catalyst binding or oxidation by catalyst [V(V) to V(IV)] as the turnover limiting step. Attempts at coupling mixed phenol/ether substrates (SI, Control F) showed decomposition of starting material with no detectable phenol-dienone product formation, indicating that oxidation likely occurs through covalent activation by V2, as opposed to outer-sphere electron-transfer from V(V). Furthermore, no homo-coupled dimeric products were observed during a standard reaction with tethered phenols, supporting a mechanism that proceeds through vanadium species (see structure II, Scheme 5), as opposed to free radical forms of oxidized 1a.

Scheme 5.

Proposed Mechanism

These control results, combined with prior findings,^12,33 support a proposed mechanism (Scheme 5) wherein phenoxide-bound substrate (I) is oxidized by two equivalents of V2. Prior work indicated the formation of a V(IV)-V(V) intermediate (II), which undergoes intersystem crossing to react via a triplet.³³ We propose that the radical on the blue ring forms first as the phenol is less hindered due to the para- disposition of the tether. The subsequent intramolecular coupling followed by substrate displacement affords bis-dienone intermediate IV and reduced catalyst species III. Bis-dienone intermediate IV undergoes a tautomerization to form the phenol portion of the final phenol-dienone product. Catalyst intermediate III is then re-oxidized in the presence of dioxygen to reform active catalyst V2. Additional mechanistic studies are ongoing, with an emphasis on identifying the nature of the species involved in the key C-C bond forming step.

In summary, a new vanadium Schiff base catalyst was developed for use in the catalytic, intramolecular coupling of tethered phenols to form phenol-dienone products. Modifications to the catalyst counterion revealed that an HFIP-coordinated V(V)-oxo catalyst was superior in mediating the coupling reaction. In total, 20 oxidatively coupled adducts were prepared in yields up to 94%, including three natural products and a morphinan alkaloid analog. This newly developed catalytic method for intramolecular phenol-dienone coupling serves as a valuable addition to the rich field of oxidative phenol C-H functionalization³⁴ and represents an effective small molecule catalytic approach to the biomimetic coupling of tethered phenols. Further studies into the mechanism and enantioselective variants of these processes are underway.