“On-Water” Synthesis of Quinazolinones and Dihydroquinazolinones Starting from o-Bromobenzonitrile

A versatile and practical “on-water” protocol was newly developed to synthesize quinazolinones using o-bromobenzonitrile as a novel starting material. Studies have found that air as well as water plays an important role in synthesis of quinazolinones. Further investigation indicated that dihydroquinazolinones can be prepared with this protocol under the protection of N2. The protocol can be extended to other substrates and various quinazolinones and dihydroquinazolinones were obtained. o-Bromobenzamide, o-aminobenzonitrile, and o-aminobenzamide were also evaluated as starting materials, and the results further proved the versatility of this protocol, especially towards dihydroquinazolinones.


Introduction
Quinazolinones and dihydroquinazolinons are important classes of nitrogen-containing heterocycles with an array of biological activities such an antitumor [1,2], anti-inflammatory [3], antibacterial [4,5], anticonvulsant [6], etc. As explicit examples, RVX-208 and balaglitazone, structurally based on quinazolin-4(3H)-one (Figure 1), are now under phase III clinical trials. The compound of RVX-208 is developed for the treatment of cardiovascular diseases and lipid metabolism disorders [7], while balaglitazone is developed for the treatment of type 2 diabetes [8]. Recently, RVX-208 was found effective to reactivate HIV-1 in latent reservoirs [9], which stimulated our interest to synthesize the compounds bearing a quinazolione core.
alternative substrate to access quinazolinones for following reasons: (1) with o-bromobenzonitrile as an alternative starting material, the scope of substrates could be substantially extended, and thus increase the variety of quinazolinones; (2) in our methodology, transforming o-bromobenzonitrile into quinazolinones in one pot can save an extra step (benzonitrile into benzamide [38,39] thus the cost can be reduced; and (3) to the best of our knowledge, the substituted o-bromobenzonitriles were usually cheaper than the corresponding o-bromobenzamides.
Meanwhile, most of these reactions using DMSO or DMA as solvent, suffered from unpleasant smells in high temperature. In addition, it is hard to remove solvent after the reaction. Langer and co-workers [36] provided a green protocol so that dihydroquinazolinons can be prepared in H2O with good yields, whereas quinazolinones can only be obtained by adding oxidant TBHP(tert-butyl hydroperoxide). In the context of green chemistry, water was widely explored as a solvent in various reactions [40][41][42][43], not only due to its low cost, non-toxicity, easy availability, and eco-benign features, but also because the theoretical and practical advantages of water substantially improved the "onwater" reaction [44][45][46]. To date, water has never been used as media to synthesize quinazolinones directly without adding any oxidant. Cognizant of these challenges, we tried to introduce water as reaction media into the synthesis of quinazolinones. To our delight, we finally explored a versatile and practical "on-water" protocol for the quinazolinone preparation from o-bromobenzonitriles with comparable yields. Surprisingly, with this protocol, a dihydroquinazolinone skeleton could also be built just under the protection of N2. Finally, we successfully synthesized 31 compounds including quinazolinones and dihydroquinazolinons, and 10 (4aa, 4ee, 4eq, 4ff, 5aa, 5ea, 5eb, 5ee, 5eq, 5fe) of them were novel compounds. Herein, we would like to present our research towards quinazolinones synthesis in detail.

Scheme 1.
In situ transformation from o-bromobenzamide to o-bromobenzamide.

Discussion and Results
Preliminary investigation started with the reaction of o-bromobenzonitrile 1a (instead of obromobenzamide) [27], benzaldehyde 2 and ammonia 3 under 100 °C in the present of CuBr, Lproline, and Cs2CO3 for 24 h. The expected product 2-phenylquinazolin-4(3H)-one 4aa was obtained (Entry 1, Table 1). The reaction conditions could further be optimized. The results collected in Table  1 showed that the reaction performed under 100 °C using Cs2CO3 as a base provided the highest yield (Entries 1-8, Table 1). The further screening of the catalyst (Entries 9-15, Table 1) showed that copper (Cu (II)) salts gave higher yields of product 4aa (Entries 12 and 14, Table 1) than Cu (I) in this case, being totally different from the reported results [11,12,25,29,31].
With the catalyst CuCl2, solvent effects were examined (Entries 16-21, Table 1) and H2O was successfully introduced. Without any oxidant, the "on-water" reaction provided compound 4aa with yield of 75% (Entry 18, Table 1), much higher than using DMF and DMA media. Further attempts to increase the yield by adding DMSO or PEG (polyethylene glycol) in water as a co-solvent failed (Entries 19-20, Table 1). Additionally, air was found to be important for the fusion of quinazolinone Meanwhile, most of these reactions using DMSO or DMA as solvent, suffered from unpleasant smells in high temperature. In addition, it is hard to remove solvent after the reaction. Langer and co-workers [36] provided a green protocol so that dihydroquinazolinons can be prepared in H 2 O with good yields, whereas quinazolinones can only be obtained by adding oxidant TBHP(tert-butyl hydroperoxide). In the context of green chemistry, water was widely explored as a solvent in various reactions [40][41][42][43], not only due to its low cost, non-toxicity, easy availability, and eco-benign features, but also because the theoretical and practical advantages of water substantially improved the "on-water" reaction [44][45][46]. To date, water has never been used as media to synthesize quinazolinones directly without adding any oxidant. Cognizant of these challenges, we tried to introduce water as reaction media into the synthesis of quinazolinones. To our delight, we finally explored a versatile and practical "on-water" protocol for the quinazolinone preparation from o-bromobenzonitriles with comparable yields. Surprisingly, with this protocol, a dihydroquinazolinone skeleton could also be built just under the protection of N 2 . Finally, we successfully synthesized 31 compounds including quinazolinones and dihydroquinazolinons, and 10 (4aa, 4ee, 4eq, 4ff, 5aa, 5ea, 5eb, 5ee, 5eq, 5fe) of them were novel compounds. Herein, we would like to present our research towards quinazolinones synthesis in detail.

Discussion and Results
Preliminary investigation started with the reaction of o-bromobenzonitrile 1a (instead of o-bromobenzamide) [27], benzaldehyde 2 and ammonia 3 under 100 • C in the present of CuBr, L-proline, and Cs 2 CO 3 for 24 h. The expected product 2-phenylquinazolin-4(3H)-one 4aa was obtained (Entry 1, Table 1). The reaction conditions could further be optimized. The results collected in Table 1 showed that the reaction performed under 100 • C using Cs 2 CO 3 as a base provided the highest yield (Entries 1-8, Table 1). The further screening of the catalyst (Entries 9-15, Table 1) showed that copper (Cu (II)) salts gave higher yields of product 4aa (Entries 12 and 14, Table 1) than Cu (I) in this case, being totally different from the reported results [11,12,25,29,31].
With the catalyst CuCl 2 , solvent effects were examined (Entries 16-21, Table 1) and H 2 O was successfully introduced. Without any oxidant, the "on-water" reaction provided compound 4aa with yield of 75% (Entry 18, Table 1), much higher than using DMF and DMA media. Further attempts to increase the yield by adding DMSO or PEG (polyethylene glycol) in water as a co-solvent failed (Entries 19-20, Table 1). Additionally, air was found to be important for the fusion of quinazolinone core, as only traces of the product formed when the transformation was conducted under the protection of N 2 (Entry 22, Table 1). Based on these results, we can conclude that air and H 2 O are vital to promote the formation of quinazolinone. Therefore, we obtained the optimal reaction conditions (CuCl 2 (0.1 mmol), Cs 2 CO 3 (2 mmol), L-proline (0.2 mmol), H 2 O (2 mL)) for the condensation of o-bromobenzonitriles, aldehydes, and aqueous ammonia toward quinazolinones, which are highlighted in Table 1 (Entry 18). core, as only traces of the product formed when the transformation was conducted under the protection of N2 (Entry 22, Table 1). Based on these results, we can conclude that air and H2O are vital to promote the formation of quinazolinone. Therefore, we obtained the optimal reaction conditions (CuCl2 (0.1 mmol), Cs2CO3 (2 mmol), L-proline (0.2 mmol), H2O (2 mL)) for the condensation of obromobenzonitriles, aldehydes, and aqueous ammonia toward quinazolinones, which are highlighted in Table 1 (Entry 18). With such an eco-friendly protocol in hand, the workup procedure was simple on account of water. Thereupon, it is necessary to investigate the scope and limitation to explore the versatility of this protocol. Before that, other o-halobenzonitriles (Entries 2-4, Table 2) were tested and the results collected in Table 2 showed that o-bromobenzonitrile is the most active substrate for this reaction. Various aryl aldehydes 2 were firstly evaluated under standard reaction conditions (Table 2). Generally, most of them were well tolerated and successfully transformed into the corresponding products with moderate to good yields. The electro-donating groups (-Me, -OMe) made a positive influence for the reaction and resulted in higher yields (4ad, 4ae, 4af, Table 2) than electrowithdrawing counterparts (4ab, 4ac, 4ai, Table 2). It is noteworthy that almost no steric effect was observed for benzaldehyde, and o-methoxybenzaldehyde 2f even gave a higher yield up to 83% (4af,  With such an eco-friendly protocol in hand, the workup procedure was simple on account of water. Thereupon, it is necessary to investigate the scope and limitation to explore the versatility of this protocol. Before that, other o-halobenzonitriles (Entries 2-4, Table 2) were tested and the results collected in Table 2 showed that o-bromobenzonitrile is the most active substrate for this reaction. Various aryl aldehydes 2 were firstly evaluated under standard reaction conditions (Table 2). Generally, most of them were well tolerated and successfully transformed into the corresponding products with moderate to good yields. The electro-donating groups (-Me, -OMe) made a positive influence for the reaction and resulted in higher yields (4ad, 4ae, 4af, Table 2) than electro-withdrawing counterparts (4ab, 4ac, 4ai, Table 2). It is noteworthy that almost no steric effect was observed for benzaldehyde, and o-methoxybenzaldehyde 2f even gave a higher yield up to 83% (4af, Table 2) than p-methoxyone (4ae, Table 2). Then we extended this protocol to substituted o-bromobenzonitriles. The electro-withdrawing group on o-bromobenzonitrile (Entries 20-23, Table 2) was beneficial to the transformation and gave higher yields of corresponding products than those electro-donating group on o-bromobenzonitrile (Entries 24-28, Table 2). Therefore, we got an excellent yield of 92% (4ef, Table 2) when 2-bromo-5-fluorobenzonitrile reacted with 2-methoxylbenzaldehyde, which further proved that electro-donating benzaldehydes, especially 2-methoxylbenzaldehyde, were more active than electro-withdrawing benzaldehydes.   Table 2). Then we extended this protocol to substituted obromobenzonitriles.
The selected substituted benzaldehydes were smoothly transformed into corresponding expected products with good to excellent yields (Table 3).
Therefore, 2-bromo-5-fluorobenzonitrile as well as 2-bromo-5-methylbenzonitrile was treated with benzaldehydes and aqueous ammonia under the the protection of N2 with standard reaction conditions (5ea-5fe, Table 3). Most of them were smoothly transformed into the expected dihydroxylproducts in good yields. Table 3. The scope and limitation for the synthesis of 2,3-dihydro-2-aryl quinazolin-4(1H)-one 5 a . According to the results and similar reactions [12,13,24,32], a possible mechanism was proposed and outlined in Scheme 2. In the presence of Cs2CO3, the coordination of L-proline with CuCl2 helped to promote the subsequent Ullmann-type reaction to form 2-amino, while oxidation from -CN to -CONH2 occurred simultaneously in the presence of CuCl2, base, and H2O, so that intermediate 2aminobenzamide was generated. Actually, 2-aminobenzamide was detected as intermediate during the reaction, andthen condensation and addition occurred on 2-aminobenzamide and benzaldehyde. 2,3-Dihydro-2-aryl quinazolin-4(1H)-one 5 was produced under the protection of N2, and when the reaction was exposed to air after the starting materials were stirred in sealed tubes at 100 °C for 12 h, oxidation occurred further to afford products 2-aryl quinazolin-4(3H)-one 4. According to the results and similar reactions [12,13,24,32], a possible mechanism was proposed and outlined in Scheme 2. In the presence of Cs 2 CO 3 , the coordination of L-proline with CuCl 2 helped to promote the subsequent Ullmann-type reaction to form 2-amino, while oxidation from -CN to -CONH 2 occurred simultaneously in the presence of CuCl 2 , base, and H 2 O, so that intermediate 2-aminobenzamide was generated. Actually, 2-aminobenzamide was detected as intermediate during the reaction, andthen condensation and addition occurred on 2-aminobenzamide and benzaldehyde. 2,3-Dihydro-2-aryl quinazolin-4(1H)-one 5 was produced under the protection of N 2 , and when the reaction was exposed to air after the starting materials were stirred in sealed tubes at 100 • C for 12 h, oxidation occurred further to afford products 2-aryl quinazolin-4(3H)-one 4. Although at the outset of this work we only expected to develop an "on-water" protocol towards quinazolinones from o-bromobenzonitrile 1a, the explicit outcome above proved the versatility of this protocol. Then other commonly used substrates were evaluated for this protocol. o-Bromobenzamide 1h was first reacted with p-methoxylbenzaldehyde and ammonia under the standard conditions. As outlined in Scheme 3, 2-(4-methoxyphenyl)-quinazolin-4(3H)-one 4ae and 2,3-dihydro-2phenylquinazolin-4(1H)-one 5ae were produced smoothly at 74% and 68%, respectively, almost having the same yields with o-bromobenzonitrile (73% for 4ae in Table 2, and 71% for 5ae in Table 3).

Scheme 3. The "on-water" reaction starting from o-bromobenzamide.
Subsequently, o-aminobenzonitrile 1i was tested and the results in Scheme 4 showed that the yield of 2-phenylquinazolin-4(3H)-one 4aa was only 43%, while o-bromobenzonitrile 1a yielded in 75% (4aa, Table 2). But, when under the protection of N2, 76% of the yield of 2,3-dihydro-2phenylquinazolin-4(1H)-one 5aa was obtained, almost the same with the result from obromobenzonitrile 1a (5aa, Table 3). Obviously, o-aminobenzonitrile 1i was less active than o- Although at the outset of this work we only expected to develop an "on-water" protocol towards quinazolinones from o-bromobenzonitrile 1a, the explicit outcome above proved the versatility of this protocol. Then other commonly used substrates were evaluated for this protocol. o-Bromobenzamide 1h was first reacted with p-methoxylbenzaldehyde and ammonia under the standard conditions. As outlined in Scheme 3, 2-(4-methoxyphenyl)-quinazolin-4(3H)-one 4ae and 2,3-dihydro-2-phenylquinazolin-4(1H)-one 5ae were produced smoothly at 74% and 68%, respectively, almost having the same yields with o-bromobenzonitrile (73% for 4ae in Table 2, and 71% for 5ae in Table 3). Although at the outset of this work we only expected to develop an "on-water" protocol towards quinazolinones from o-bromobenzonitrile 1a, the explicit outcome above proved the versatility of this protocol. Then other commonly used substrates were evaluated for this protocol. o-Bromobenzamide 1h was first reacted with p-methoxylbenzaldehyde and ammonia under the standard conditions. As outlined in Scheme 3, 2-(4-methoxyphenyl)-quinazolin-4(3H)-one 4ae and 2,3-dihydro-2phenylquinazolin-4(1H)-one 5ae were produced smoothly at 74% and 68%, respectively, almost having the same yields with o-bromobenzonitrile (73% for 4ae in Table 2, and 71% for 5ae in Table 3).
Molecules 2018, 23, x FOR PEER REVIEW 7 of 12 bromobenzonitrile 1a for the synthesis of quinazolin-4(3H)-one, which explained why 2,3-dihydroquinazolin-4(1H)-ones were the only products listed in the literature reported by Langer and coworkers [36], wherein oxidant TBHP had to be added in addition to help the formation of quinazolin-4(3H)-one.

Conclusions
In summary, we have newly developed a versatile and practical "on-water" protocol towards compounds containing a quinazolinone core, and o-bromobenzonitrile was explored as an alternative This interesting finding prompted us to undertake the reaction starting from o-aminobenzamide 1j immediately, as shown in Scheme 5. Unexpectedly, oxidized product 2-phenylquinazolin-4(3H)-one 4aa failed to produce, while 2,3-dihydro-2-phenylquinazolin-4(1H)-one 5aa yielded up to 95% without the protection of N 2 . It implies that o-aminobenzamide 1j tends to be transformed into 2,3-dihydro-quinazolin-4(1H)-ones 5 with this protocol.
Molecules 2018, 23, x FOR PEER REVIEW 7 of 12 bromobenzonitrile 1a for the synthesis of quinazolin-4(3H)-one, which explained why 2,3-dihydroquinazolin-4(1H)-ones were the only products listed in the literature reported by Langer and coworkers [36], wherein oxidant TBHP had to be added in addition to help the formation of quinazolin-4(3H)-one.

Conclusions
In summary, we have newly developed a versatile and practical "on-water" protocol towards compounds containing a quinazolinone core, and o-bromobenzonitrile was explored as an alternative With these results from the extended substrates, we found the newly developed "on-water" protocol was much more flexible for the construction of 2,3-dihydro-quinazolin-4(1H)-ones 5 than its oxidized products quinazolin-4(3H)-ones 4. In addition, for the preparation of quinazolin-4(3H)-ones 4, the priority of the commonly used substrates is: o-bromobenzonitrile 1a or o-bromobenzamide 1h > o-aminobenzonitrile 1i > o-aminobenzamide 1j.

Conclusions
In summary, we have newly developed a versatile and practical "on-water" protocol towards compounds containing a quinazolinone core, and o-bromobenzonitrile was explored as an alternative starting material. Water and air were found to be of importance for the formation of oxidized products quinazolin-4(3H)-ones 4, and various aryl aldehydes as well as several substituted o-bromobenzonitriles were successfully applied to this protocol and provided the expected quinazolin-4(3H)-ones 4. Moreover, we found that 2,3-dihydro-quinazolin-4(1H)-ones 5 could also be obtained when the reaction was carried out under the protection of N 2 , and the subsequent scope investigation resulted in the successful synthesis of various 2,3-dihydro-quinazolin-4(1H)-ones 5. Therefore, we can selectively produce the compounds 2,3-dihydro-quinazolin-4(1H)-ones 5 or their oxidized products 4 from o-bromobenzonitrile by controlling air. In addition, o-bromobenzonitrile, o-aminobenzonitrile, and o-aminobenzamide were evaluated and all of them could be transformed into 2,3-dihydro-quinazolin-4(1H)-ones 5 with good to excellent yields, but o-bromobenzonitrile or o-bromobenzamide was proved to be the best substrate for the synthesis of the oxidized products quinazolin-4(3H)-ones compared with o-aminobenzonitrile and o-aminobenzamide. With these new findings in mind, further work involving the synthesis of RVX-208 and related structural modifications are ongoing in our group.

General
All chemicals were purchased from commercial suppliers Shanghai Energy Chemical Co., Ltd. (Shanghai, China), Adamas Reagent, Ltd. (Shanghai, China), TCI Industry Co., Ltd. (Shanghai, China), without further purification. All reactions were monitored by TLC (thin-layer chromatography) which was performed on GF254 silica gel glass plates (Qingdao Haiyang Chemical Co. Ltd., Qingdao, Shandong, China). Column chromatography was performed with silica gel (200-300 mesh). All unknown compounds were structurally verified by 1 H-NMR, 13 C-NMR and MS, and 1 H-, and 13 C-NMR spectra were recorded on a Bruker Advance drx 400 spectrometer (Bruker Bioscience, Billerica, MA, USA) operating at 400 MHz and 100 MHz, respectively. The chemical shifts were reported in ppm and the coupling constant in Hz. Mass Spectrometry analysed for the known compounds by Waters HPLC/ZQ 4000 Thermo Fisher Scientific (Waltham, MA, USA).

General Procedure for the Synthesis of 2-Phenylquinazolin-4(3H)-one (4aa)
To a mixture of 2-Bromobenzonitrile (183.4 mg, 1 mmol), benzaldehyde (210.5 mg, 2 mmol), CuCl 2 (17.2 mg, 0.1 mmol), Cs 2 CO 3 (652.2 mg, 2 mmol), and L-proline (23.2 mg, 0.2 mmol) in H 2 O (2 mL) was added 27% aqueous ammonia (1 mL) in a tube under air atmosphere. Then the tube was sealed, and the mixture was stirred at 100 • C for 12 h. Next, the tube was opened to air and the mixture was stirred at 100 • C for another 12 h. After being cooled to room temperature, the resulting mixture was quenched with NH 4 Cl solution and extracted with ethyl acetate. The combined organic layer was washed with brine, and then dried over anhydrous Na 2 SO 4 . The solvent was evaporated under reduced pressure and the crude product was purified by chromatography on silica-gel to afford 2-phenylquinazolin-4(3H)-one (4aa) in 75% isolated yield. 1   into a tube and stirred. Remove the air inside the tube under the reduced pressure and flush with N 2 , and protected the starting materials under N 2 . 27 % aqueous ammonia (1 mL) was added into the reaction mixture. The tube was then sealed and the mixture stirred at 100 • C for 24 h. After cooling to room temperature, the resulting mixture was quenched with NH 4 Cl solution and extracted with ethyl acetate. The combined organic layers were washed with brine and then dried over anhydrous Na 2 SO 4 . The solvent was evaporated under reduced pressure and the crude product was purified by chromatography on silica-gel to afford 2-phenyl-2,3-dihydroquinazolin-4(1H)-one(5aa) in 74% isolated yield. 1  o-Aminobenzanitrile (119.8 mg, 1 mmol), benzaldehyde (217.8 mg, 2 mmol), CuCl 2 (17.6 mg, 0.1 mmol), Cs 2 CO 3 (652.3 mg, 2 mmol), and L-proline 23.6 mg, 0.2 mmol) in H 2 O (2 mL) was added in a 5 mL reaction bottle. Then the mixture was stirred at 100 • C for 48 h. After cooling to room temperature, the resulting mixture was quenched with NH 4 Cl solution and extracted with ethyl acetate. The combined organic layers were washed with brine and then dried over anhydrous Na 2 SO 4 . The solvent was evaporated under reduced pressure and the crude product was purified by chromatography on silica-gel to afford 2-phenylquinazolin-4(3H)-one (4aa) in 43% isolated yield.
Experimental procedures and analytical data of all compounds ( 1 H NMR and 13 C NMR), copy of the 1 H NMR and 13 C NMR and data are available in the Supplementary Materials.