Synthesis and Spectral Characterization of 4,7-Dichloro-6-nitroquinazoline

: Afatinib is a 4-anilinoquinazoline tyrosine kinase inhibitor (TKI) in the form of a dimaleate salt which is indicated for the treatment of locally advanced or metastatic non-small cell lung cancer (NSCLC). The most scalable route for the synthesis of this drug was reported in two Boehringer Ingelheim patents, in which the title compound, 4,7-dichloro-6-nitroquinazoline ( IV ), is an important intermediate. Compound IV is also present in a number of synthetic pathways for various 4,7-disubstituted quinazoline derivatives displaying high therapeutic potential. However, no detailed characterization of this popular compound has been reported, possibly due to its high instability. In this paper, IV was prepared in an overall yield of 56.1% by a 3-step process (condensation, nitration, and chlorination) from 2-amino-4-chlorobenzoic acid ( I ). The target compound has been for the ﬁrst time fully characterized by melting point, mass-spectrometry, FT-IR, 1 H-NMR and 13 C-NMR spectroscopies. A methanol as

In that synthetic procedure, the compound 4,7-dichloro-6-nitroquinazoline (IV, CAS Registry number 162012-71-7) is a highly reactive intermediate. However, it was used in situ and thus no characterization was available. Similarly, in other synthetic routes of other bioactive quinazoline derivatives reported by SciFinder and Reaxys, compound IV Molbank 2020, 2020, M1134 2 of 7 was also synthesized and used directly in the next step without characterization , or only partially characterized by 1 H-NMR [37]. The most detailed characterization of IV includes a melting point analysis, 1 H-NMR and 13 C-NMR spectroscopies, and an elemental analysis [38]. However, it appears that the reported NMR data for IV in this paper are close to those of the starting material (7-chloro-6-nitroquinazolin-4(3H)-one, III), indicating the possibility of a hydrolysis reaction which converts IV back to the starting material. In addition, the given elemental analysis did not present the percentage of each element, which offers no clarification over the identity of this compound. In this paper, compound IV has been for the first time fully characterized by melting point, mass-spectrometry, FT-IR, 1 H-NMR and 13 C-NMR spectroscopies.

Results and Discussion
The target compound IV was prepared in three steps from 2-amino-4-chlorobenzoic acid (I) in an overall yield of 56.1% (Scheme 1). The synthetic procedures were based on references [30,33,39] with some changes.
Molbank 2020, 2020, x 5 of 5 other bioactive quinazoline derivatives reported by SciFinder and Reaxys, compound IV was also synthesized and used directly in the next step without characterization , or only partially characterized by 1 H-NMR [37]. The most detailed characterization of IV includes a melting point analysis, 1 H-NMR and 13 C-NMR spectroscopies, and an elemental analysis [38]. However, it appears that the reported NMR data for IV in this paper are close to those of the starting material (7-chloro-6-nitroquinazolin-4(3H)-one, III), indicating the possibility of a hydrolysis reaction which converts IV back to the starting material. In addition, the given elemental analysis did not present the percentage of each element, which offers no clarification over the identity of this compound. In this paper, compound IV has been for the first time fully characterized by melting point, massspectrometry, FT-IR, 1 H-NMR and 13 C-NMR spectroscopies.

Results and Discussion
The target compound IV was prepared in three steps from 2-amino-4-chlorobenzoic acid (I) in an overall yield of 56.1% (Scheme 1). The synthetic procedures were based on references [30,33,39] with some changes. Compound IV is an imidoyl halide (also known as imidyl or iminochlorides), a group of highly reactive organic compounds widely used as synthetic tools to produce a variety of compounds. It is often not necessary to isolate imidoyl halides because in situ generation and subsequent reactions can afford the desired derivatives in high yields [40][41][42]. In fact, our experimental observations demonstrate that IV is extremely sensitive to moisture and can be readily hydrolyzed giving back the starting material III. This phenomenon could explain the reason why the NMR-data in the aforementioned article [38] are so close to those of the starting material III. These data are presented in Tables 1 and 2.  Compound IV is an imidoyl halide (also known as imidyl or iminochlorides), a group of highly reactive organic compounds widely used as synthetic tools to produce a variety of compounds. It is often not necessary to isolate imidoyl halides because in situ generation and subsequent reactions can afford the desired derivatives in high yields [40][41][42]. In fact, our experimental observations demonstrate that IV is extremely sensitive to moisture and can be readily hydrolyzed giving back the starting material III. This phenomenon could explain the reason why the NMR-data in the aforementioned article [38] are so close to those of the starting material III. These data are presented in Tables 1 and 2.  The data in Table 1 indicate that III is capable of tautomerism between the lactim (-C(OH)=N-) and lactam (-C(=O)-NH-) forms, displaying a huge difference in chemical shifts between the -OH and the -NH groups (12.79 ppm vs. 3.30 ppm). In the latter form, the -NH-signal is actually overlapped with the water signal if DMSO-d 6 is used and may not be seen in a proton spectrum. As the reported proton chemical shifts for compound IV in [38] are similar to those of compound III in [30], we suspect that they are of the same chemical compound (Table 1). Indeed, the 1 H-NMR values for compound IV in [38] do not match up with those of either our data or another paper [37], displaying large chemical shifts' differences of 0.96 ppm, 0.44 ppm, and 0.43 ppm, respectively, for the aromatic hydrogens in DMSO-d 6 [37,38]. We would not expect such large discrepancies in chemical shifts for the same compound using the same NMR solvent (DMSO-d 6 ). When comparing our 1 H-NMR data with those of [37], the differences are −0.38 ppm, 0.05 ppm, and 0.02 ppm, respectively, which indicate a high level of spectral similarities. These slight deviations could potentially originate from the fact that two different NMR solvents (DMSO-d 6 vs. CDCl 3 ) were used [44]. On the other hand, when examining the 13 C-NMR data, we found strikingly similar values between IV of reference [38] and the starting material III (∆δ = 0-0.4 ppm, Table 2). These reported values are different from our 13 C-NMR data of IV (∆δ = 0.4-5.4 ppm, Table 2). Altogether, both the 1 H-NMR and 13 C-NMR data combined indicate that the reported NMR data in [38] could potentially be that of the starting material III, and not the desired compound IV. One possible explanation could be the hydrolysis of the highly reactive compound IV back into the starting material III, a phenomenon we have observed in our laboratory.
Another interesting fact is that, when measuring mass spectrometry, we not only detected the peak of compound IV, but also its methoxy form (compound V), which possibly occurs due to alcoholysis by methanol (the solvent in MS). We show this in Scheme 2. *: The signal is overlapped with water signal at 3.33 ppm in DMSO-d6. ** NA: not available. The data in Table 1 indicate that III is capable of tautomerism between the lactim (-C(OH)=N-) and lactam (-C(=O)-NH-) forms, displaying a huge difference in chemical shifts between the -OH and the -NH groups (12.79 ppm vs. 3.30 ppm). In the latter form, the -NH-signal is actually overlapped with the water signal if DMSO-d6 is used and may not be seen in a proton spectrum. As the reported proton chemical shifts for compound IV in [38] are similar to those of compound III in [30], we suspect that they are of the same chemical compound (Table 1). Indeed, the 1 H-NMR values for compound IV in [38] do not match up with those of either our data or another paper [37], displaying large chemical shifts' differences of 0.96 ppm, 0.44 ppm, and 0.43 ppm, respectively, for the aromatic hydrogens in DMSO-d6 [37,38]. We would not expect such large discrepancies in chemical shifts for the same compound using the same NMR solvent (DMSO-d6). When comparing our 1 H-NMR data with those of [37], the differences are −0.38 ppm, 0.05 ppm, and 0.02 ppm, respectively, which indicate a high level of spectral similarities. These slight deviations could potentially originate from the fact that two different NMR solvents (DMSO-d6 vs. CDCl3) were used [44]. On the other hand, when examining the 13 C-NMR data, we found strikingly similar values between IV of reference [38] and the starting material III (∆δ = 0-0.4 ppm, Table 2). These reported values are different from our 13 C-NMR data of IV (∆δ = 0.4-5.4 ppm, Table 2). Altogether, both the 1 H-NMR and 13 C-NMR data combined indicate that the reported NMR data in [38] could potentially be that of the starting material III, and not the desired compound IV. One possible explanation could be the hydrolysis of the highly reactive compound IV back into the starting material III, a phenomenon we have observed in our laboratory.
Another interesting fact is that, when measuring mass spectrometry, we not only detected the peak of compound IV, but also its methoxy form (compound V), which possibly occurs due to alcoholysis by methanol (the solvent in MS). We show this in Scheme 2. On the whole, in the synthesis of IV, our control of anhydrous reaction conditions and the appropriate work-up procedures have made it feasible to synthesize and characterize this highly reactive compound for the first time. These data may be useful for further investigations in the synthesis process improvement of afatinib, its analogs and other biologically active quinazoline based On the whole, in the synthesis of IV, our control of anhydrous reaction conditions and the appropriate work-up procedures have made it feasible to synthesize and characterize this highly reactive compound for the first time. These data may be useful for further investigations in the synthesis process improvement of afatinib, its analogs and other biologically active quinazoline based compounds. All the mass, FT-IR, 1 H-NMR and 13 C-NMR spectra are presented in the Supplementary Material File.
The melting point was determined using a SRS EZ-Melt apparatus (Stanford Research Systems, Sunnyvale, CA, USA) and is uncorrected. MS was performed at a EVOQ Qube™ (Bruker, Billerica, MA, USA) or an LTQ Orbitrap XL™ (Thermo Scientific, Waltham, MA, USA) system. FT-IR spectra were recorded by a Perkin Elmer (Waltham, MA, USA) or Shimadzu (Kyoto, Japan) spectrometer. 1 Hand 13 C-NMR spectra were acquired with a 500 MHz Ascent spectrometer (Bruker, Billerica, MA, USA) using acetone-d 6 , DMSO-d 6 , or CDCl 3 as the solvent. The reaction mixtures were monitored, and the purity of all products was checked by thin-layer chromatography (TLC) on silica gel 60 F 254 plates (Merck, Darmstadt, Germany).

Preparation of 7-Chloroquinazolin-4(3H)-one (II)
The synthetic procedure for compound II was based on the method described in [39] with some modifications. A mixture of 2-amino-4-chlorobenzoic acid (I, 34.30 g, 0.20 mol, 1 equiv.), formamide (125.0 mL, 2.77 mol, 14 equiv.) was aerated with nitrogen, then stirred and refluxed for 1.5 h at 160 • C (the reaction was monitored by TLC with a 9:1 DCM/methanol mixture as eluent). The reaction mixture was cooled to 80 • C, then 500 mL of water was added. The mixture was cooled to −5 • C in 1 h and filtered. The precipitate was washed with water and dried at 60 • C to afford 7-chloroquinazolin-4(3H)-one (II) as a light brown solid (29.71 g, 82.3%), which was used for the next step without further purification. The synthetic procedure for compound III was based on [30] with some modifications. Compound II (10.83 g, 0.060 mol) and sulfuric acid (120 mL) were added into a 500 mL two-neck round bottom flask. The mixture was cooled to 0 • C on ice and stirred until dissolution. Fuming nitric acid (120 mL) was slowly added to the mixture at 0 • C, and the mixture was stirred for 1.5 h at 30 • C. After completion of reaction as indicated by TLC, 10% NaOH solution was slowly added to the reaction mixture until a precipitate was formed (pH~7). The mixture was then filtered to furnish a light yellow solid (III). The compound was purified by redissolving in HCl (5M) and then neutralizing this solution with NaOH (10%) to pH 6-7 to furnish the pure precipitate which was then filtered, washed with water and dried at 60 • C to furnish 7-chloro-6-nitroquinazolin-4(3H)-one (III) as a light yellow solid (11.