Rapid and catalyst free synthesis of new bis(benzo[g]chromene) and bis(pyrano[3,2-c]chromene) derivatives and optimization of reaction conditions using response surface methodology

4,4′-(1,4-phenylene)bis(2-(alkylamino)-3-nitro-4H-benzo[g]chromene-5,10-dione) and 4,4′-(1,4-phenylene)bis(2-(alkylamino)-3-nitropyrano[3,2-c]chromen-5(4H)-one) derivatives are synthesized by a one-pot, multi-component reaction of N-alkyl-1-(methylthio)-2-nitroethenamine (derived from the reaction of various amines and 1,1-bis(methylthio)-2-nitroethene) with terephthalaldehyde or isophthalaldehyde, and 2-hydroxy-1,4-naphthoquinone or 4-hydroxycoumarin in EtOH/H2O (85 : 15) as the solvent at 89 °C. Response surface methodology (RSM) is used to investigate the effect of reaction temperature and water content of aqueous ethanol on the product yields and reaction time. The notable features of this work are the optimization of reaction conditions with minimal experiments, absence of catalyst, good yields, simple work-up and the non-chromatographic purification of products.


Introduction
Response surface methodology (RSM), is one of the most commonly used experimental designs for optimizing process conditions. [1][2][3][4][5][6][7] Traditionally, optimization has been carried out by monitoring the inuence of one factor at a time on an experimental response. 8 RSM is a more economical method because a small number of experiments are carried out for monitoring the interaction of the independent variables in the response. In conventional optimization, the increase in the number of experiments necessary to perform the research, leads to an increase in time, expense, and the consumption of reagents and materials for experiments. 8,9 Hence, response surface methodology (RSM) may be considered as an efficient way to deal with the limitations of the conventional method. 10 In the recent years, several studies used successfully RSM for optimizing reaction conditions such as optimization of reaction parameters for the synthesis of chromium methionine complex (Xiao et al., 2013), 5 optimization of the synthesis of highly functionalized pyrimido [1,2-b]indazoles via 6-endo-dig cyclization (Roopan et al., 2016), 11 and optimization of the reaction conditions for the synthesis of 1-(2-aminoethyl)-2-imidazolidone (Cai et al., 2017). 12 Multi-component reactions (MCRs) represent a powerful and efficient method for construction of complex molecules with biological potential. 13 They offer remarkable advantages like convergence, operational simplicity, facile automation, reduction in the number of workups, extraction and purication processes. 14 Therefore, the design of novel MCRs has attracted great attention from synthetic organic chemists. [15][16][17][18][19][20] Chromenes and their fused analogues are a group of biologically active molecules occurring extensively in nature with a wide range of molecular modications. 21 Among them benzochromenes are one of the privileged scaffolds with a medicinal pharmacophore such as antibacterial, 22 antimicrobial, 23 antitumor, 24 and anti-inammatory. [25][26][27][28] Also pyranochromenes are an important and essential category of chromenes and have extensive bioactivities, such as anti-HIV, anti-tuberculosis, 29 antitumor, 30 anti-cancer, anti-anaphylactic, 31 and anticoagulant 32 activities.

Result and discussion
In the current study, we introduce an efficient one-pot, multicomponent reaction of various amines 1, 1,1-bis(methylthio)-2-nitroethene 2, terephthalaldehyde or isophthalaldehyde 3, Response surface methodology (RSM) using central composite design (CCD) with ve replicates at the center point was applied to optimize the reaction conditions. We used the reaction of N-methyl-1-(methylthio)-2-nitroethenamine (2 mmol), terephthalaldehyde (1 mmol), and 2-hydroxy-1,4naphthoquinone (2 mmol) for the preparation of compound 6b as a model. The two independent variables (temperature (A) and water content of aqueous ethanol (B)), levels (Table 1), and the results obtained aer running the experiments are represented in Table 2. The two responses were analyzed: yield of product (R 1 ) and reaction time (R 2 ).
Optimum conditions with respect to yield, purity of product and reaction time were as follow: temperature 89 C, water content of aqueous ethanol 15%. Verication experiments, carried out at the predicted conditions showed values reasonably close to those predicted and further conrmed the adequacy of predicted models.
On the basis of the CCD, the cubic model relationship between the experimental yield (R 1 ) and the process variables (temperature (A) and water content of aqueous ethanol (B)) and the quadratic model between the reaction time (R 2 ) and the process variables, in coded units are obtained from eqn (1) and (2) respectively. R 1 ¼ 22.80 + 22.63A À 14.85B À 13.00AB + 4.60A 2 + 1.60B 2 À 11.65A 2 B À 7.13AB 2 (1) where R 1 and R 2 represent the experimental yield and reaction time respectively, then A (temperature) and B (water content of aqueous ethanol) are the coded variables in the reaction. Fig. 1 and 2 show the linear correlation between the actual and predicted yield and time, respectively. According to the gures, the predicted yield is consistent with the experimental yield and the model can be used to predict the yield of product successfully.
The ANOVA for response surface cubic and quadratic models are reported in Tables 3 and 4. According to the Table 3, the model F-value of 46.05 implies the cubic model is signicant. There is only a 0.03% chance that a "Model F-Value" this large could occur due to noise. Values of "Prob > F" less than 0.0500 indicate model terms are signicant. In this case A, B, AB, A 2 B are signicant model terms. The "Lack of Fit F-value" of 0.00 implies the Lack of Fit is not signicant relative to the pure error. There is a 100.00% chance that a "Lack of Fit F-value" this large could occur due to noise. Non-signicant Lack of Fit is good and we want the model to t. The "Pred. R-Squared" of 0.9761 is in reasonable agreement with the "Adj. R-Squared" of  There is only a 0.06% chance that a "Model F-Value" this large could occur due to noise. The "Pred. R-Squared" of 0.5210 is not as close to the "Adj. R-Squared" of 0.8845 as one might normally expect. This may indicate a large block effect. Also "Adeq. Precision" measures the signal to noise ratio. A ratio greater than 4 is desirable. Our ratio of 12.884 indicates an adequate signal. The effects of temperature and water content of aqueous ethanol on the total reaction yield, and the total reaction time are shown in Fig. 3 and 4, respectively. The total yield varied from 0% to 84%. As temperature increased, the yield increased and the reaction time decreased. The changes in yield versus the water content of aqueous ethanol is minor compared to that of temperature. As the water content of aqueous ethanol decreased, the yield increased and the reaction time decreased. The highest yield (84%) was obtained at 89 C and at 15% water. The lowest yield (0%) were at 36 C, 85% water, and at 25 C, 50% water.
Aer obtaining the optimized reaction conditions, we investigated the scope of this reaction by varying the structure of the amines 1, phthalaldehydes 3, and CH acids (4 and 5). The reaction proceeds cleanly under the same reaction conditions to afford corresponding fused heterocyclic derivatives 6 and 7 (Schemes 3 and 4) in 75-91% yields. The results are shown in Table 5. The reaction did not work, when the reaction was performed using ortho-phthalaldehyde, and benzylamine. The steric hindrance makes it less reactive toward reaction.
Structural elucidation of the newly synthesized compounds was accomplished using their IR, mass, 1 H NMR, and 13 C NMR spectroscopy (see the ESI †). Compounds 6 and 7 have two stereogenic centers, and therefore two diastereoisomers are expected. For examples the two diastereoisomers of 6a is shown in Scheme 2. The 1 H and 13 C NMR spectra of the products are consistant with the presence of two diastereoisomers. We were not able to separate compounds 6 and 7 in pure form. However, their NMR data can be extracted from the mixture of the two diastereoisomers in a nearly 50 : 50 ratio (based on integration of methine protons). The 1 H and 13 C NMR spectra of two diastereoisomers are similar except for the alkyl amine and methine groups. Also we did not obtain the good NMR spectra  For examples the IR spectrum of 6b indicated absorption bands due to the NH stretching (3613 cm À1 ) as well as bands at 1667, 1467 and 1363, 1245, and 1192 cm À1 due to the C]O, NO 2 , C-N, and C-O groups. The 1 H NMR spectrum of 6b showed multiplets for four CH 3 groups (d 3.15-3.23 ppm), two singlets for four CH groups (d 5.23 and 5.28 ppm), multiplets for the aromatic region (d 7.01-8.04 ppm), and multiplets for four NH groups (d 10.23-10.35 ppm). The 1 H-decoupled 13 C NMR spectrum of 6b showed three signals for CH 3 and CH groups (d 29.9, 37.4, 38.0 ppm), and signals at 108.5, 177.6, 182.8, 183.5, and 184.7 ppm, which were assigned C-NO 2 and C]O groups, respectively. The mass spectrum of this structure was in accordance with the proposed structure.