Transition metal-free synthesis of 2-aryl quinazolines via alcohol dehydrogenation

We report here a transition metal-free synthesis of quinazoline derivatives starting from 2-aminobenzyl alcohols and aryl amides via an alcohol dehydrogenation strategy promoted by potassium tertiary butoxide. The control experiments are carried out to identify the reaction intermediates and the role of the K + ion in the reaction. The DFT calculations unveil the reaction mechanism, with special focus on the rate determining state. The present method tolerates a variety of functional groups providing easy access to diversely substituted quinazolines.

In recent years, several procedures have been developed for the synthesis of quinazoline derivatives. Most of the methods are based on either oxidative condensation or coupling reactions [18][19][20][21][22][23][24][25][26][27]. However, they are associated with severe drawbacks, such as the use of chemically unstable o-aminobenzaldehyde as the reactant [28], more than stoichiometric amounts of hazardous oxidants [29], generation of a large quantity of hazardous waste [30], etc. Thus, an efficient method, using benign chemicals under eco-friendly mild reaction conditions and without producing much waste, is well appreciated.
In this context, acceptorless dehydrogenative coupling (ADC) reactions have emerged as an efficient tool for synthesizing diverse heterocyclic compounds [31,32], from relatively inexpensive and readily available starting materials [33,34]. The only by-products generated in this type of reaction are hydrogen and water [35][36][37], which makes the strategy environmentally benign. One prominent example in this category is the dehydrogenation of alcohols followed by coupling with a suitable reagent [38]. The above strategy has been extended for the synthesis of substituted quinazoline derivatives (see Scheme 1) by several research groups. For example, Paul and coworkers disclosed the synthesis of aryl quinazoline by nickel-catalyzed alcohol dehydrogenation followed by condensation with 2-aminobenzylamine [39]. In similar reports, Balaraman and co-workers revealed the synthesis of quinazoline using 2-aminobenzyl alcohol and amides as reactants and the manganese pincer complex as a catalyst in the presence of 0.4 equivalents of potassium-tert-butoxide (KO t Bu) [40].
In general, 2-aminobenzyl alcohols are found to undergo alcohol dehydrogenation to form 2-amino benzaldehyde as an intermediate that coupled with amide [40] or nitrile [41][42][43], resulting in the formation of quinazoline derivatives. In another report, Li and coworkers showed that iron and phenanthroline catalyst systems combined with CsOH⋅H 2 O as a base could be used for the same transformation [44].
The use of heavy transition metals in organic reactions have posed some serious limitations as many of these metals are in general hazardous to the environment, expensive, and scarce in nature [72,73]. Similarly, the ligands used in these reactions are generally difficult to prepare and expensive. Moreover, metal and ligand contamination especially in pharmaceutical products are closely regulated. Hence, transition metal-free protocols are gaining tremendous importance in overcoming these challenges [74][75][76]. The base-mediated β-alkylation of alcohols in aerobic conditions is well known [77]. In line with heterogeneous bases have been demonstrated as catalysts for dehydrogenative processes [78][79][80][81], the use of potassium-tert-butoxide in dehydrogenation is not new and was reported by Grubbs and coworkers in the silylation of C-H bonds in aromatic heterocycles [82]. Later Yu and his group disclosed potassium tert-butoxide-promoted acceptorless dehydrogenation of N-heterocycles [83]. Recently potassium-tert-butoxide mediated direct synthesis of amides from alcohol was reported by Fang and co-workers via an MPV-type hydrogen transfer process [84]. Thus, to overcome the aforementioned challenges involved in transition metal catalysis and to meet the requirement of improved transition metal free protocols in heterocyclic chemistry [85][86][87][88][89][90][91][92][93][94], we report here a simple and straightforward method for synthesizing 2-arylquinazoline derivatives by dehydrogenative coupling between 2-aminobenzyl alcohol and an amide in the presence of K t OBu. The experimental procedure is very simple and straightforward, does not involve any transition metal catalyst [95]. A mixture of 2-aminobenzyl alcohol and benzamide in t-amyl alcohol was stirred at 100 • C in the presence of K t OBu for a required period of time (TLC). Standard workup and purification by column chromatography afforded the product.

Experimental section
General procedures for the synthesis of quinazoline. A mixture of aryl amide (0.5 mmol), 2-aminobenzyl alcohol (0.75 mmol), and KO t Bu (1.0 mmol) in 2 mL t-amyl alcohol was heated at 100 • C under argon atmosphere for 16 h in a preheated heating block. After completion of the reaction, the reaction vessel was cooled to room temperature and diluted with 10 mL ethyl acetate. The reaction mixture was then filtered using celite. The filtrate was dried using a rotary evaporator, volatile impurities were removed under vacuum, and further purification of the product had been carried out by column chromatography using silica gel as stationary phase and hexane and ethyl acetate as eluent. The purified products are characterized by IR and 1 H NMR and 13 C NMR spectroscopy.
2-(p-tolyl)quinazoline (3a,    Computational Details: DFT static calculations were performed with the Gaussian16 set of programs [96], using the BP86 functional of Becke and Perdew [97][98][99], including corrections due to dispersion through the Grimme's method (GD3 keyword in Gaussian16) [100,101]. The electronic configuration of the molecular systems was described with the double-ζ basis set with the polarization of Ahlrichs for main-group atoms (def2-SVP keyword in Gaussian) [102]. The geometry optimizations were performed without symmetry constraints, and analytical frequency calculations confirmed the character of the located stationary points. These frequencies were used to calculate unscaled zero-point energies (ZPEs) as well. Energies at 373.15 K were obtained by single-point calculations on the optimized geometries with the M06-D3 functional [103,104]. and the triple-ζ basis set def2-TZVPP and by estimating solvent effects with and estimation of solvent effects with the universal solvation model SMD of Cramer and Truhlar for amyl alcohol [105]. The reported Gibbs energies in this work include electronic energies obtained at the M06-D3/def2-TZVPP (smd)//BP86-D3/def2-SVP level of theory corrected with zero-point energies, thermal corrections and entropy effects computed with the BP86-D3/def2-SVP level.

Results and discussion
To optimize the reaction conditions, the reaction was studied with various reaction parameters such as reaction temperature, time, solvents, and the amount of base used, and the results are summarized in Table 1. Similarly, the ratio of the reactant was also standardized and reported in the supporting information. It was found that the maximum yield of the product was obtained with one equivalent of benzamide, 1.5 equivalents of 2-aminobenzyl alcohol, and two equivalents of KO t Bu as a base in t-amyl alcohol at 100 • C for 16 h (entry 2, Table 1).
The solvent screening showed that toluene was slightly less effective than t-amyl alcohol and provided a relatively low yield (entries 1 and 2, Table 1). The conversion was dramatically influenced by polar aprotic solvents like DMF (entry 5, Table 1) and a more polar solvent like water (entry 6, Table 1). Time vs conversion plot (Fig. S2, SI) showed a minimum of 16 h reaction period was required for the completion of the reaction (entry 2, Table 1). Additionally, it had also been observed that to consume the limiting reagent (benzamide) completely, an excess (1.5 equivalents) of 2-aminobenzyl alcohol is required (Table 4, SI).
Furthermore, the reaction yield was significantly reduced when KO t Bu was replaced by a weaker base (entries 9-12, Table 1). The reaction was not initiated with potassium carbonate as a base (entry 11, Table 1). Interestingly, during the optimization, we noticed that temperature plays a crucial role in this reaction. By decreasing the temperature to 80 • C, the conversion decreased drastically (entry 8, Table 3 Control Experiments. a Standard condition: 2-aminobenzyl alcohol (0.75 mmol), amide (0.5 mmol), base (1 mmol), solvent (2 ml), at 100 • C under Ar for 16 h. b Yield is determined by 1 H NMR spectroscopy using 1,4 dimethoxybenzene as the internal standard. c Yield refers to that of a pure product characterized by 1 H NMR. Table 1). At room temperature, the response was futile (entry 7, Table 1). Several structurally varied substituted aryl and aliphatic amide underwent couplings with reaction intermediate generated in-situ from 2-aminobenzyl alcohol by this procedure to produce the corresponding quinazolines derivatives. The results are reported in Table 2.
Both aromatic and heterocyclic amides (3k, Table 2) participated efficiently in this reaction to form 2-substituted quinazolines (Table 2) irrespective of the electronic nature of the substituent present. However, ortho-substituted benzamides (3c, 3d, Table 2) react sluggishly, and even after 24 h of reaction, only 72% of corresponding products are isolated. The meta-substituted nitro benzamide also furnished a comparatively low yield (3j, Table 2). In general, the substituents at m and p position in aryl amides, e.g., 4-Me (3a,  Table 2) and 3-OMe (3i, Table 2), did not pose any difficulty to produce the corresponding quinazoline in high yields. Significantly, the reaction proceeded with aliphatic amide (3l, Table 2) without any interference. Moreover, substituted 2-aminobenzyl alcohol (3n, 3o, 3p, Table 2) and heteroaryl benzyl alcohol (3 m, Table 2) participated in the reaction effectively to furnish the corresponding product in good yield. This indicates the moderately good substrate scope for this reaction.
To understand the reaction mechanism and identify the reaction intermediate, several control experiments (Table 3) were performed at different time intervals and the results were critically analyzed. During the control reaction, in the absence of benzamide, we observed the formation of 39% of 2-aminobenzyladehyde as the product (entry 2, Table 3).
In another experiment, 4% of 2-aminobenzaldehyde was identified when the reaction was quenched at the intermediate stage (entry 4, Table 3), clearly indicating the formation of 2-amino benzaldehyde as an intermediate. Surprisingly, it has been observed that the NaO t Bu was not effective in initiating the reaction (entry 5, Table 3). A similar trend was also noticed when we used NaOH (entry 12, Table 1) as a base when compared to KOH (entry 9, Table 1). we envisaged that the potassium ion plays a significant role in this reaction. To unravel this hypothesis, metal-ion trapping experiments with 18-crown-6 have been performed. In the presence of 18-crown-6, the progress of the reaction was affected remarkably (entry 6, Table 3), which undoubtedly indicates the crucial role of potassium ions in the reaction pathway. Anyway, it is not ensured if it acts catalytically. Furthermore, it has been observed that the presence of free radical quencher, TEMPO in excess, did not have much impact on the overall reaction progress (entry 3, Table 3). Therefore, the
Berkessel et al. reported base-catalyzed hydrogenation of aromatic ketones and proposed a mechanism in which the reaction proceeds through a cyclic transition state involving carbonyl C=O, H -H, and O-K in KO t Bu [106]. Since dehydrogenation is just the reverse of hydrogenation, a similar mechanism may also be operative for this potassium tertiary butoxide-promoted alcohol dehydrogenation. Based on the literature reports [77,106] and the results of the control experiments in Table 3, we believed that density functional theory (DFT) calculations could shed light on the reaction mechanism. Among the most plausible strategies, including alcohol dehydrogenation [83], hydrogenation without a metal catalyst [106], and borrowing hydrogen [71,107]. Fig. 2 includes the preferred mechanism calculated at the M06-D3/def2-TZVPP(smd)//BP86-D3/def2-SVP level of theory, following the metal-free transformation [108].
The mechanism in Fig. 2 begins with the proton transfer from the alcohol group of 2-aminobenzyl alcohol to KO t Bu with the amino group that works as a proton shuttle. This step is kinetically feasible by overcoming an energy barrier of 13.0 kcal/mol through TS 2→3 (see Fig. 3a). Therefore, from 3 the reaction proceeds through the six-membered ring transition state TS 3→4 with an energy barrier of 26.8 kcal/mol such as in the dehydrogenation reported by Berkessel et al. (see Fig. 3b).
This step facilitates the formation of molecular hydrogen as well as the corresponding aldehyde 5, which is nearly isoenergetic to the analogous initial alcohol 1, simply being 0.9 kcal/mol higher in energy. At this point of the reaction, a unit of KO t Bu removes one proton from the amide group of the other reagent. As a result, an amide anion stabilized by potassium is formed and t BuOH is released into the solution. Thereafter, the anion reacts with the aldehyde 5, to form an imine intermediate in several steps. As reported in the literature [109]. the imine formation is step-wise: first the carbinolamine formation, followed by the dehydration step. In our case study, the formation of the carbinolamine is not favored by the protic solvent, with the amyl alcohol acting as a proton shuttle that transfers a proton from the amine group to the carbonyl [110,111]. On the other hand, the potassium cation can activate the carbonyl groups and bring the reactants closer. In detail, here from 5 the reaction proceeds with the formation of complex 6 between aldehyde and the amide anion with a stabilization of 4.7 kcal/mol, considering that we omit the corrections of entropy and standard state of 1 M concentration in solution [112]. Subsequently, carbinolamine 7 is formed overcoming an energy barrier of 12.5 kcal/mol after the nucleophilic attack of the amide anion to the aldehyde carbonyl. Furthermore, the imine is formed not through dehydration, but by elimination of a KOH molecule. At the beginning, the rather unstable complex 8 with the protic solvent is formed and the remaining amino proton is transferred to the potassium alkoxide moieties forming a zwitterionic intermediate through TS 8→9 overcoming an energy barrier of 5.5 kcal/mol in the subsequent step. Then, KOH is removed through TS 9→10 (ΔG ‡ = 17.4 kcal/mol) and the imine 11 is formed after releasing KOH••• t BuOH into the solution. The newly formed imine is deprotonated by KOH or KO t Bu, forming complexes 12. Then, the deprotonated amino group reacts with the aldehyde moiety forming a six-membered ring of potassium amino alkoxide 13 through TS 12→13, overcoming energy barriers of 10.6 and 9.5 kcal/mol with R = t Bu and Finally, the quinazoline product 18 is formed by releasing a KOH molecule, overcoming an energy barrier of only 5.0 kcal/mol corresponding to TS 16→17. Overall, the step defined by TS 9→10 results to be the rate determining step [113], actually the rate-determining state (rds) [114] of the whole reaction mechanism, with an overall kinetic cost of 34.9 kcal/mol [115][116][117]. However, it is also necessary to consider above all the other determining step, which is TS 3→4, the reason derives from seeing that if we replace KO t Bu with NaO t Bu as a reagent, the rds described as TS 16→17 remains almost isoenergetic, and even being 0.4 kcal/mol lower, while the step defined by TS 9→10 does get significantly worse by 2.2 kcal/mol. To give more reliability to this trend, the TS 16→17 was also studied. Although it would not be decisive for anything, it is also seen how its kinetic cost increases by 2.6 kcal/mol with the sodium ion. This leads us to conclude that sodium as a countercation of the base has a worse behavior than potassium. Moreover, it gives us feedback by making us see that the rds would not be TS 9→10 but TS 3→4, thus being a kinetic cost of 30.9 kcal/mol, also more in agreement with the experimental temperature of 100 • C. The reason why the TS 9→10 barrier is overestimated is that even though it was attempted, more than one t BuOH molecule would have to be involved.
An alternative mechanisms has also been proposed in Fig. 4 starting with imine 11 where one molecule of 2-aminobenzyl alcohol hydrogenated the imine group similarly to the pyridine-mediated alcohol oxidation reported by Namitharan and coworkers [107]. The 2-aminobenzyl alcohol is deprotonated to form the corresponding potassium alkoxide. Next, the potassium alkoxide forms complex 19 with imine 11 required to overcome TS 19→20 with an overall kinetic cost of 36.9 kcal/mol, which discards this alternative mechanism. On the other hand, another mechanism starts with the formation of the potassium 2-aminobenzyl alkoxide 22, with a favorable energy gain of − 3.7 kcal/mol. The alkoxide then forms the adduct 23 with the amide. Next, overcoming an energy barrier of 26.0 kcal/mol there is the H-transfer that leads to intermediate 24, with the fundamental role of the potassium cation to stabilize the negative charge of any of the oxygen atoms involved. Through the rotational movement of the K-O-C -N dihedral angle, complex 24 isomerizes into complex 25 to facilitate the proton transfer from the amino group to the carbonyl. However, although the aldehyde is activated by potassium, the condensation of aldehyde with the amine group results to be expensive in terms of kinetics (ΔG ‡ = 31.9 kcal/mol) and it leads to the formation of species 26 (see Fig. 3d). The next step, the dehydration step [37,38,118], results to be expensive too even if assisted by a solvent molecule [119][120][121]. First, intermediate 27 is formed and, then the remaining amino proton is transferred through TS 27→28 (ΔG ‡ = 13.3 kcal/mol) to the hydroxyl group meanwhile the latter one is acting as a leaving group. Anyway, the latter steps lead to an overall kinetic cost of up to 51.2 kcal/mol. Since both of the latter mechanisms discussed are higher in energy with respect to the one displayed in Fig. 2, they were discarded.

Conclusions
In conclusion, we have developed a methodology for the synthesis of substituted quinazolines under transition metal-free reaction conditions, and omitting oxidations working under argon atmosphere [122]. Controlled experiments were carried out to elucidate the reaction mechanism. The role of K + ions was established by trapping with 18-crown-6-ether. The DFT calculations not only unveiled the reaction mechanism but also justify the reaction conditions, laying the groundwork for improving them even further by describing the most kinetically demanding steps. The present method offers operational simplicity and general applicability to synthesize a wide range of 2-substituted quinazolines derivatives (including alkyl, aryl, and heterocyclic), with good to excellent yields of products. To the best of our knowledge, we are unaware of any report for quinazoline synthesis starting from 2-aminobenzyl alcohol and amides as substrates without using a transition metal as a catalyst.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability
A summary is collected in the supporting information file and then more details are available upon request.