Facile Synthesis of 2H-Benzo[h]Chromenes via an Arylamine-Catalyzed Mannich Cyclization Cascade Reaction

A simple arylamine-catalyzed Mannich-cyclization cascade reaction was developed for facile synthesis of substituted 2H-benzo[h]chromenes. The notable feature of the process included the efficient generation of ortho-quinone methides (o-QMs) catalyzed by a simple aniline. The mild reaction conditions allowed for a broad spectrum of 1- and 2-naphthols and trans-cinnamaldehydes to engage in the cascade sequence with high efficiency.

Along with the classical approaches, significant efforts have been devoted to developing more efficient catalytic methods, largely with transition metals [2,3]. In contrast, there are only a handful of reports about the synthesis of the 2H-benzo[h]chromenes. The methods are dominated by the acid-catalyzed condensation of naphthols with propargylic alcohol by p-toluenesulfonic acid [11,12], acidic alumina [13], indium trichloride [14], and montmorillonite K-10 [15]. Nucleophilic addition dehydration reactions of chromanones with Grignard [16] or organolithium [17] reagents provide an alternative. Cycloaddition reactions of naphthols with simple α, β-unsaturated aldehydes are also attractive strategies. However, the reactions are performed under harsh (refluxing) reaction conditions [18][19][20][21][22]. New methods with mild reaction conditions will streamline the synthesis of 2H-benzo[h]chromenes, which can tolerate broad functional groups. Herein, we wish to document a simple arylamine catalyzed formation of the molecular architectures under mild reaction conditions with a broad substrate scope (Scheme 1). A new reactivity involving the aniline-catalyzed in situ formation of ortho-quinone methide (o-QM) in a catalytic fashion is proposed. The reactivity is harnessed for developing a synthetically powerful Mannich-Diels-Alder cascade manifold for the construction of 2H-benzo[h]chromenes.

Design Plan
o-QM, a highly reactive intermediate originally harnessed in nature for var logical purposes [23], has been slowly recognized by synthetic organic chemists in synthesis [24][25][26][27][28][29]. Given the high reactivity of o-QMs, their in situ formation from responding stable aromatic precursors by temporarily masking the "O" is gene ried out [24,25]. Therefore, an additional activation step is often required. In add synthesis of the precursors is not trivial. We propose a new, more efficient cata proach for the in situ generation of o-QMs from readily accessible simple α, β-uns aldehydes (enals) (Scheme 1). We envision that the direct functionalization of hyde group of enals 1 with an amine catalyst gives imine 4. The reaction of 4 with thol 2 via a Mannich-type leads to product 5, which serves as a precursor for a su o-QM (6) formation by extrusion of the amine catalyst. Finally, an intramolecula tion reaction delivers benzopyrans, driven by re-aromatization.
Although the proposed work looks straightforward on paper, there are sev barriers to overcome for achieving the desired cascade transformations. First, th dition reaction (Mannich) of 1 with 1-naphthols 2 may compete with the pos conjuage addition process, which has been observed in chiral pyrrolidine-catal chael cyclization [30]. This suggests that the amine catalysts play key roles in th the products formed. To minimize the 1,4-conjuage addition process, weakly nucl less-hindered aromatic amines may be the choice of promoters for achieving the 1,2-addition. It is believed that the aniline-derived imine 4 favors the 1,2-over 1,4reaction, because its less-electrophilic and more-hindered features reduce the p of 1,4-addition, while the less-steric hindrance enhances the 1,2-addition accessib ond, achieving an unprecedented catalytic process for the formation of o-QMs appealing but formidably challenging. Precedent studies relied on the preformati precursors using a stoichiometric amount of amine, and the generation of o-QMs light or acid activation [31,32]. We contemplated that the better leaving tendency o may enable the process in a catalytic manner. The nucleophilicity and leaving ten amine need to be balanced to catalyze this cascade process effectively. Due to t leaving ability of arylamines compared with aliphatic amines, we successfully de aniline-promoted cyclization replacement cascade reactions of 2-hydroxycinn hydes with various carbonic nucleophiles for the synthesis of dihydrobenzopyra

Design Plan
o-QM, a highly reactive intermediate originally harnessed in nature for various biological purposes [23], has been slowly recognized by synthetic organic chemists in organic synthesis [24][25][26][27][28][29]. Given the high reactivity of o-QMs, their in situ formation from the corresponding stable aromatic precursors by temporarily masking the "O" is generally carried out [24,25]. Therefore, an additional activation step is often required. In addition, the synthesis of the precursors is not trivial. We propose a new, more efficient catalytic approach for the in situ generation of o-QMs from readily accessible simple α, β-unsaturated aldehydes (enals) (Scheme 1). We envision that the direct functionalization of the aldehyde group of enals 1 with an amine catalyst gives imine 4. The reaction of 4 with 1-naphthol 2 via a Mannich-type leads to product 5, which serves as a precursor for a subsequent o-QM (6) formation by extrusion of the amine catalyst. Finally, an intramolecular cyclization reaction delivers benzopyrans, driven by re-aromatization.
Although the proposed work looks straightforward on paper, there are several key barriers to overcome for achieving the desired cascade transformations. First, the 1,2addition reaction (Mannich) of 1 with 1-naphthols 2 may compete with the possible 1,4conjuage addition process, which has been observed in chiral pyrrolidine-catalyzed Michael cyclization [30]. This suggests that the amine catalysts play key roles in the fate of the products formed. To minimize the 1,4-conjuage addition process, weakly nucleophilic, less-hindered aromatic amines may be the choice of promoters for achieving the desired 1,2-addition. It is believed that the aniline-derived imine 4 favors the 1,2-over 1,4-addition reaction, because its less-electrophilic and more-hindered features reduce the possibility of 1,4-addition, while the less-steric hindrance enhances the 1,2-addition accessibility. Second, achieving an unprecedented catalytic process for the formation of o-QMs is highly appealing but formidably challenging. Precedent studies relied on the preformation of the precursors using a stoichiometric amount of amine, and the generation of o-QMs requires light or acid activation [31,32]. We contemplated that the better leaving tendency of aniline may enable the process in a catalytic manner. The nucleophilicity and leaving tendency of amine need to be balanced to catalyze this cascade process effectively. Due to the better leaving ability of arylamines compared with aliphatic amines, we successfully developed aniline-promoted cyclization replacement cascade reactions of 2-hydroxycinnamaldehydes with various carbonic nucleophiles for the synthesis of dihydrobenzopyrans [33].

Exploration
To test the possibility of in situ-generated o-QMs promoted by aniline, followed by an intramolecular Diels-Alder reaction, a reaction between trans-cinnamaldehyde (1a) and 1-naphthol (2a) was conducted in the presence of a catalytic amount of an arylamine (20 mol%; Table 1). Gladly, the reaction with 20 mol% of I proceeded to give the desired product 2-phenyl-2H-benzo[h]chromene (3a), albeit a low yield of 10% after 48 h. Encouraged by the results, other anilines containing electron-donating or -withdrawing substituents were probed. 2-Methyoxyaniline (III) was much better than 4-methyoxyaniline (II) (entry 2 vs. 3) for this cascade process, whereas an electron-deficient group like 4nitroaniline (IV) was less effective (entry 4). Naphthalen-1-amine (V) and a secondary aniline 1,2,3,4-tetrahydroquinoline (VI) were employed, showing an extremely low catalytic potency (entries 5 and 6). Then, we thought about whether ortho-hydrogen-bonding donor groups in the catalyst could accelerate this reaction. Thus, the 2-aminophenol (VII), 2-aminobenzoic acid (VIII) and o-phenylenediamine (IX) aniline analogs were studied (entries 7-9). IX gave the highest yield of 56% (entry 9). Therefore, we chose IX as the choice for further optimization. To test the possibility of in situ-generated o-QMs promoted by aniline, followed by an intramolecular Diels-Alder reaction, a reaction between trans-cinnamaldehyde (1a) and 1-naphthol (2a) was conducted in the presence of a catalytic amount of an arylamine (20 mol%; Table 1). Gladly, the reaction with 20 mol% of I proceeded to give the desired product 2-phenyl-2H-benzo[h]chromene (3a), albeit a low yield of 10% after 48 h. Encouraged by the results, other anilines containing electron-donating or -withdrawing substituents were probed. 2-Methyoxyaniline (III) was much better than 4-methyoxyaniline (II) (entry 2 vs. 3) for this cascade process, whereas an electron-deficient group like 4-nitroaniline (IV) was less effective (entry 4). Naphthalen-1-amine (V) and a secondary aniline 1,2,3,4-tetrahydroquinoline (VI) were employed, showing an extremely low catalytic potency (entries 5 and 6). Then, we thought about whether ortho-hydrogen-bonding donor groups in the catalyst could accelerate this reaction. Thus, the 2-aminophenol (VII), 2aminobenzoic acid (VIII) and o-phenylenediamine (IX) aniline analogs were studied (entries 7-9). IX gave the highest yield of 56% (entry 9). Therefore, we chose IX as the choice for further optimization. The solvent effect on the reaction was probed next (Table 1, entries 18-24). Et 2 O and THF produced low-yielding products (entries 18 and 19). The same happened with polar solvents MeOH and MeCN (entries 20 and 21). This reaction was favored by nonpolar aprotic reaction media (entries 9 and 22-24). Further studies were carried out to optimize the Mannich cyclization cascade reaction in CH 2 Cl 2 with catalyst IX (Table 1). Even though the reaction time was prolonged and the temperature was raised to 40 • C, no further improvement was attended (entries 10 and 11). Increasing the amount of 2a to 1.5 equiv. gave better results (entry 12). When 1-naphthol (2a) was used as a limiting reactant, significant increase in yield (65%) was observed (entry 13). Furthermore, higher temperature, it not only reduced the reaction time but also enhanced the efficiency (entry 14). Finally, when the reaction was performed in dry DCM under Ar atmosphere, the yield was further improved, suggesting a negative effect of water in the reaction (87%, entry 15). A decreased catalyst loading lowered its efficiency (entry 16). Interestingly, p-phenylenediamine (X) gave a very low yield, which indicates that ortho-diamine plays a critical role in catalysis (entry 17). Therefore, the optimal reaction conditions for the formation of 3a involved the usage of 0.3 mmol of 1a, 0.2 mmol of 2a and 20 mol% of IX in 1 mL of dry DCM under Ar.

Reaction Scope
An investigation was carried out to probe the scope of the reactions of the substituted trans-cinnamaldehyde 1 with 1-naphthol (2) (Scheme 2). It seems the process served as a general approach for the production of structurally diverse 2H-benzo[h]chromenes. The reaction was less sensitive to electronic and steric effects. The enals with electronwithdrawing groups such as nitro, chloro, floro and trifloromethyl all gave rise to good yields (3b-d, 3g). A similar trend was observed for trans-cinnamaldehydes bearing electrondonating groups, which furnished 2H-benzo[h]chromenes 3h-l in high yields (3h-l). The steric demand produced a minimal impact as well (3b, 3f, 3g and 3i-l). Notably, the mild protocol tolerated various functional groups, such as MeO, hydroxyl, acetoxy, NMe 2 , etc. Next, we probed the structural variations of 1-napthnols (Scheme 1). The protocol could tolerate electron-withdrawing groups such as Cl, Br, NO 2 , OMe, NHBoc and NHAc to give the corresponding products 3m-r in good yields. Finally, the heteroaromatic enals, including furan, pyridine and quinolone, were examined. They participated in the process to smoothly deliver the desired products 3s-u with high efficiency.

Materials and Methods
All commercially available reagents were purchased from Sigma-Aldrich, TCI, Alfa, Ambeed in the US and were used without further purification. All reactions were carried out in dried flasks. The progress of the reactions was monitored by analytical thin-layer chromatography (TLC) on Whatman silica gel plates with a fluorescence F254 indicator. Merck 60 silica gel was used for chromatography. 1 H (300 MHz) and 13 C (75 MHz) NMR spectra were recorded on a Bruker Avance 300. When deuterated chloroform (CDCl3) was used, residue chloroform was used as an internal reference. When deuterated dimethyl sulfoxide (DMSO) was used, residue dimethyl sulfoxide was used as an internal reference. Data for 1 H NMR were reported as follows: chemical shift (ppm) and multiplicity (s = singlet, d = doublet, t = triplet, q = quartet and m = multiplet). Chemical shift for 13 C NMR was reported as ppm.

Materials and Methods
All commercially available reagents were purchased from Sigma-Aldrich, TCI, Alfa, Ambeed in the US and were used without further purification. All reactions were carried out in dried flasks. The progress of the reactions was monitored by analytical thin-layer chromatography (TLC) on Whatman silica gel plates with a fluorescence F254 indicator. Merck 60 silica gel was used for chromatography. 1 H (300 MHz) and 13 C (75 MHz) NMR spectra were recorded on a Bruker Avance 300 (Supplementary Materials). When deuterated chloroform (CDCl 3 ) was used, residue chloroform was used as an internal reference. When deuterated dimethyl sulfoxide (DMSO) was used, residue dimethyl sulfoxide was used as an internal reference. Data for 1 H NMR were reported as follows: chemical shift (ppm) and multiplicity (s = singlet, d = doublet, t = triplet, q = quartet and m = multiplet). Chemical shift for 13 C NMR was reported as ppm.