Synthesis of hexahydroquinoline (HHQ) derivatives using ZrOCl₂·8H₂O as a potential green catalyst and optimization of reaction conditions using design of experiment (DOE)

Abstract

In this investigation, hexahydroquinoline (HHQ) derivatives were synthesized via a one-pot reaction using dimedone, β-ketoester, ammonium acetate, and different aryl aldehydes. ZrOCl₂·8H₂O was used as a potential green catalyst, it is a commercially available solid material, with low toxicity, low cost and high activity, and it is easy to handle. The reaction conditions were optimized using response surface methodology (Central Composite Design (CCD)) with three replicates at a central point. Optimization showed that the optimum reaction temperature and amount of catalyst are 83.75 °C and 0.15 mol%, respectively. The lower reaction yields at temperatures higher than 83.75 °C are related to the formation of a new crystalline phase of ZrOCl₂·8H₂O. The fitted quadratic polynomial model applied to the experimental yield could well predict the experimental reaction yield. Ecofriendly reaction conditions, easy workup procedure, the reusability of the catalyst, short reaction times and high yields are some of the advantages of this work.

---

Introduction

Multicomponent reactions (MCRs) are one-pot reactions with more than two starting materials reacting together, where most of the atoms of the starting materials are present in the target molecule.¹ MCRs are atom economic, effective, convergent, and show a high bond-forming-index (BFI) (several non-hydrogen atom bonds are formed in a one-pot reaction).² So, MCRs are often more useful than conventional sequential multistep syntheses.

One of the most important and conventional MCRs is the synthesis of dihydropyridine (DHP) which is attributed to Arthur Hantzsch who discovered it in 1881.³ Due to the versatility and the general stability of the products, the Hantzsch method has remained the most common method for the synthesis of 1,4-dihydropyridines. Firstly, dihydropyridines were discovered to be an active part of nicotinamide adenine dinucleotide (NADH), the essential reducing coenzyme in animals. Finally, nifedipine (a DHP derivative) came to market as a calcium channel modulating agent.⁴ DHPs were primarily developed as cardiovascular agents, but they are also vasodilator, antihypertensive, bronchodilator, antiatherosclerotic, hepatoprotective, antitumor, antimutagenic, geroprotective, and antidiabetic agents.⁵ Their widespread pharmacological properties have interested researchers to find new derivatives which are more effective, selective, stable, and perhaps with different modes of action.⁶ Modification of the DHP ring and its substituents is a combinatorial chemistry that makes it possible to prepare a large number of compounds which can be analyzed using structure activity relationship to design better DHPs.^7,8 One of the possible structural scaffold modifications of DHP is incorporating fused rings to the DHPs, this leads to hexahydroquinoline (HHQ) derivatives, which were successfully explored by Safak’s group.⁹

HHQs clearly show the remarkable potential of novel dihydropyridine derivatives as sources of valuable drug candidates. HHQ derivatives possess a variety of biological activities, such as vasodilatory, bronchodilatory, antiatherosclerotic, antitumor, geroprotective, hepatoprotective, and antidiabetic properties.^10,11

Without using any catalyst, the transformation of starting materials to HHQ derivatives requires long reaction times, harsh reaction conditions, and large quantities of organic solvents and commonly gives low yields. So, various catalysts such as Lewis acids,^12,13 bases,^14,15 salts,¹⁶ and ionic liquids^17,18 have been used to solve the mentioned problems. Although some of these are successful, most of them are expensive, toxic, and difficult to separate from the reaction medium. Therefore, it is important to apply an effective catalyst that does not have the above mentioned problems. So, we have reported a clean, ecofriendly, facile, and rapid solvent-free reaction for the synthesis of HHQ derivatives in the presence of ZrOCl₂·8H₂O. This is a commercially available solid material, with low toxicity (LD₅₀ ZrOCl₂·8H₂O oral rate = 2950 mg kg⁻¹), low cost, high activity and is easy to handle. The zirconium(IV) compounds are ecofriendly and potential green catalysts or reagents which are used in many organic reactions under mild conditions with excellent yields.^18,19 It should be noted that, many chemists optimize their reactions using one variable at a time (OVAT). Optimization by the OVAT method is valid only when the variables do not have any interaction with each other.²⁰ In this study, the optimization of the reaction conditions was done by design of experiments (DOE).

Experimental

Materials

All of the chemicals were purchased from Merck and Fluka Chemical Companies. The products were identified using ¹H NMR, ¹³C NMR, mass analysis and melting points as well as IR spectra. The corresponding spectral data have been reported in the ESI.† The ¹H and ¹³C NMR (500 MHz) were recorded on a Bruker Avance DPX-400 FT-NMR spectrometer (δ in ppm). Mass spectra were recorded on Shimadzu GC MS-QP 1000 EX 85 apparatus. Melting points were recorded on Büchi B-545 apparatus in open capillary tubes. Infrared spectra of the products were recorded by Perkin Elmer PE-1600-FTIR. Progress of the reactions was monitored by TLC using silica gel SIL G/UV 254 plates.

General procedure for the preparation of HHQs

ZrOCl₂·8H₂O (0.0451 g, 14 mol%) as a catalyst, was added to a mixture of dimedone (0.28 g, 2 mmol), aryl aldehyde (2 mmol), ethyl acetoacetate (0.260, 2 mmol) and ammonium acetate (0.185 g, 2.4 mmol) in a test tube, then the resulting mixture was firstly stirred magnetically, and after solidification of the reaction mixture with a small rod, at 85 °C. After the completion of the reaction, as monitored by TLC, the reaction mixture was cooled to room temperature. Then, ethyl acetate (25 mL) was added, stirred and refluxed for 3 minutes. The solid catalyst was collected using filter paper and separated from the solution of the product and the remaining starting materials. The crude product was purified by recrystallization from ethyl acetate as a less toxic class III solvent.²¹ The reaction is shown in the Fig. 1.

Fig. 1 Preparation of hexahydroquinoline under optimum conditions using ZrOCl₂·8H₂O as a clean and ecofriendly catalyst.

Results and discussion

Firstly, to find the optimum reaction conditions for the synthesis of the HHQ derivatives, a reaction of aryl aldehydes, dimedone (5,5-dimethylcyclohexane-1,3-dione), ethyl acetoacetate and ammonium acetate was selected as a representative reaction. The reaction was carried out in various solvents with ZrOCl₂·8H₂O as a catalyst to investigate the effect of solvent on the reaction time. In addition, the reaction was carried out with a catalyst in solvent-free conditions. From a time and yield point of view, the reaction in solvent-free conditions is more efficient than in the presence of solvents. The results are given in Table 1.

Table 1 Time and yield of the model reaction with and without solvent in the presence of ZrOCl₂·8H₂O as a catalyst

Entry	Solvent	Time (min)	Yield^a (%)
			50 °C
a Isolated yield.
1	—	5	96
2	EtOAc	25	93
3	CH2Cl2	60	95
4	H2O	60	50
5	n-Hexane	60	80
6	Acetonitrile	15	95
7	EtOH	60	86

---

Therefore, the solvent-free method is more efficient (entry 1) and can be selected for the representative reaction. On the other hand, Lewis acid catalysts such as ZrOCl₂·8H₂O, ZrO₂, and FeCl₃ were applied to investigate the effect of various catalysts on the reaction (Table 2). As shown in Table 2, using ZrOCl₂·8H₂O (10 mol%) as a catalyst, leads to a higher yield and shorter reaction time.

Table 2 Effect of different catalysts on the time and yield of the reaction

Entry	Catalyst	Amount of catalyst (mol%)	Time (min)	Yield^a (%)
a Isolated yield.
1	ZrOCl2·8H2O	10	5	96
2	ZrO2	10	15	90
3	FeCl3	10	20	85

---

Statistical analysis and the model fitting

The Central Composite Design (CCD) as a response surface method, with three replicates at the central point was employed to fit the experimental data to a polynomial model. Two main factors that can affect the yield of the reaction are temperature (X₁) and the amount of catalyst (X₂). These variables were coded to three levels of +1, 0, and −1. The levels of the variables and the corresponding response values (reaction yield) are shown in Table 3.

Table 3 Levels of the experimental variables and the corresponding response values of the CCD

Runs	Independent variables				Dependent variable
	X1 (temperature)		X2 (amount of catalyst)
	Coded levels	Actual levels	Coded levels	Actual levels	Yield
1	0	75.00	0	0.10	92.00
2	+1	100.00	0	0.10	85.00
3	−1	50.00	−1	0.05	67.00
4	0	75.00	0	0.10	91.00
5	0	75.00	−1	0.05	88.00
6	−1	50.00	+1	0.15	73.00
7	−1	50.00	0	0.10	67.00
8	+1	100.00	+1	0.15	93.00
9	+1	100.00	−1	0.05	73.00
10	0	75.00	0	0.10	93.00
11	0	75.00	+1	0.15	95.00

---

The reaction yield was used as the dependent variable to investigate the effect of the independent variables on the reaction yield; all of the runs (reactions) in Table 3 were stopped after 2 minutes. Analysis of variance (ANOVA) shows that the quadratic model can well predict the experimental data (Table 4). The p-values show that the model is significant from a statistical point of view. The smaller the p-value, the more significant the term. The coefficient of determination (R²) for the yield is 0.9813 with the derived model, which demonstrates that the theoretical values are in good agreement with the experimental data. The polynomial response surface model for yield based on actual values are derived from the experimental design is as follows:

Y(yield) = −59.37 + 3.73X₁ − 102.10X₂ + 2.8X₁X₂ − 0.025X₁² + 10.53X₂² (1)

Table 4 Analysis of variance for the response surface quadratic model for yield

Source	p-val. prob. > F
Model (yield)	0.0003
X1	0.0004
X2	0.0015
X1X2	0.0223
X1^2	<0.0001
X2^2	0.9852

---

It can be seen from the coefficient in eqn (1) that X₂ (amount of catalyst) has a greater impact on the reaction yield than X₁ (temperature) and that there is a considerable interaction between X₁ and X₂. Generally, the terms that have minus and plus sign have negative and positive effects on the yield, respectively. The magnitude of the effect of X is related to the value of the coefficients in Y. Fig. 2 represents the contour and three dimensional maps of the reaction yield versus the temperature and the amount of catalyst. As it is shown in Fig. 2, by increasing the amount of catalyst, the reaction yield increases, but as the temperature increases, a maximum point can be seen in the reaction yield.

Fig. 2 The contour and three dimensional maps of the reaction yield versus the temperature and the amount of the catalyst.

This behavior can be related to the dehydration/decomposition of ZrOCl₂·8H₂O at temperatures above 85 °C. From the results obtained using differential thermal analysis (DTA) and thermogravimetric analysis (TGA),^22–24 a broad endothermic feature is evident in the DSC curve for zirconyl chloride octahydrate up to about 55 °C, which reflects the removal of weakly held water molecules (lattice waters) from the crystal between 25 and 55 °C. The integrity of the tetranuclear zirconyl cations appears to be unaffected within this temperature range.²⁵ Above 55 °C, a prominent exothermic peak centered at about 73 °C and a minor exothermic maximum at about 85 °C have been observed, which indicate an increase in the degree of structural ordering (crystallization). TGA data has revealed the removal of the third lattice and one of the four coordinatively bound water molecules, respectively, at 73 and 90 °C. Thus, the removal of these two water molecules results in the formation of a new crystalline phase. For the formation of the tetrahydrate a reduction of the Zr coordination number from eight to seven has been proposed.²⁶ Two well-separated endothermic peaks have been observed above 110 °C (125 and 175 °C), which indicate a significant reduction in crystallinity of the compound (melting). Melting might be initiated by the removal of chloride counterions above 100 °C.²⁴

Based on the above discussion, yield reduction at temperatures above 85 °C can be related to the removal of the third lattice water molecule and one of the four coordinatively bound water molecules, new crystalline phase formation, and the change of the Zr coordination number ion from eight to seven. It can be deduced that tetranuclear zirconyl cations are the actual catalytic species. On the other hand, this finding is in good agreement with our previous research that the maximum yield of the reaction in the presence of ZrOCl₂·8H₂O as a catalyst was obtained at 60 watt of microwave power.²⁷

To investigate the connectivity between the decomposition/dehydration of ZrOCl₂·8H₂O and its catalytic activity, 0.075 g of catalyst was heated for 30 minutes at different temperatures. The catalyst was weighed before and after heating. Then the reaction was carried out at room temperature with the decomposed/dehydrated catalyst. Infrared spectra, the weight of the catalyst before and after heating, and the reaction yields are provided in Table 5.

Table 5 The effect of temperature on the dehydration/decomposition of the ZrOCl₂·8H₂O

Heating temperature (°C)	Weight (before)	Weight (after)	Yield (%)	Image	IR spectra
50	0.075	0.0616	70
75	0.075	0.0556	96
100	0.075	0.0495	90
125	0.075	0.0407	76
150	0.075	0.0347	65
200	0.075	0.0341	51
250	0.075	0.0338	32

---

Optimization of the reaction conditions and validation of the model

Optimization of eqn (1) for the constrained problem showed that the optimum reaction conditions are at X₁ = 83.75 and X₂ = 0.15. At these optimum conditions, the corresponding maximum yield is 99.21%. The validity of the model for predicting the reaction yield was tested using the optimum condition. The predicted and experimental optimum responses are shown in Table 6.

Table 6 Predicted and experimental value of responses at the optimum condition

Optimum	Optimum variables		Optimum response
	X1	X2	Time (min)	Yield (%)
Predicted	83.75	0.15	1	99.21
Experimental	85.00	0.15	1	97.00

---

A mean value of 97.00% (N = 3) with a deviation of 2.27% for yield was obtained from the experimental results, showing that they are in good agreement with the predicted responses. This shows that one can predict the experimental yield using eqn (1) to an acceptable deviation. In order to investigate the effect of ZrOCl₂·8H₂O as a catalyst on the reaction yield, the optimum conditions were repeated without the catalyst (X₁ = 85.00 and X₂ = 0.15). The result shows that the presence of ZrOCl₂·8H₂O is a key factor to decrease the reaction time to reach a specified yield (Table 7).

Table 7 Checking optimum conditions without the catalyst

	Variables		Response
	X1	X2	Time (min)	Yield (%)
With catalyst	85.00	0.14	<1	96
Without catalyst	85.00	0.14	>25	94

---

Using optimal conditions

After the optimization of the reaction conditions, the efficiency and applicability of the method were studied using the reaction of dimedone, ethyl acetoacetate and ammonium acetate with different aryl aldehyde derivatives in the presence of ZrOCl₂·8H₂O. The results are given in Table 8. As Table 8 shows, hexahydroquinonine derivatives could be obtained in high to excellent yields (75.41–95.82%) within short reaction times (40–180 s). Benzaldehyde derivatives with electron-releasing and electron-withdrawing substituents on the 2′, 3′, 4′, 5′ and 6′ positions and heterocyclic aldehydes were successfully tested under these reaction conditions (Table 8, compounds 1–15). Using the optimum conditions, a lower reaction time and an almost equal reaction yield can be obtained compared to those reported in the literature.^28–37

Table 8 Times, yields, and melting points for the products using optimal conditions. A = time (s), B = yield^a (%), C = melting point (°C)

a Yield refers to isolated product. Molar ratio: dimedone, ethyl acetoacetate, aldehyde, ammonium acetate (1 : 1 : 1 : 1.2). Reaction conditions: temperature (85 °C), catalyst (0.14 mole).
A	40	60	50	60
B	97	81	85	91
C	219–222 (ref. 35)	205–207 (ref. 28)	257–259 (ref. 33)	233–235 (ref. 29)
A	90	150	180	40
B	80	90	89	86
C	231–234 (ref. 36)	244–245 (ref. 33)	243–245 (ref. 35)	228–230 (ref. 32)
A	40	80	60	180
B	84	98	90	80
C	264–265 (ref. 35)	204–206 (ref. 34)	205–207 (ref. 31)	251–253 (ref. 37)
A	60	100	60
B	87	79	86
C	232–233 (ref. 28)	242–243 (ref. 30)	246–248 (ref. 33)

---

Proposed mechanism

A possible mechanism is given in Fig. 3 which is supported by the literature.³⁴ ZrOCl₂·8H₂O is a Lewis acid catalyst that catalyzes Knoevenagel type coupling of aldehydes with active methylene compounds (direction 2 and 5) and Michael type addition reactions (direction 3 and 7). To investigate the possible mechanism and the proper role of ZrOCl₂·8H₂O, an infrared (IR) technique was used. After 20 seconds at the optimum reaction conditions, two intermediates were detected and separated by the plate. The possible structures of these intermediates were identified and characterized using FT-IR. Details of the collected data are summarized in the ESI.†

Fig. 3 The possible mechanism for the synthesis of hexahydroquinoline derivatives catalyzed by ZrOCl₂·8H₂O.

Regeneration of the catalyst

The ZrOCl₂·8H₂O catalyst is a yellowish solid. Firstly, 0.15 moles of ZrOCl₂·8H₂O were added to the reaction mixture at the optimum reaction temperature (85 °C), after the reaction was complete (40 seconds), 25 ml of ethyl acetate was added to the reaction mixture at 50 °C and was mixed for 5 minutes. All of the unreacted reagents and product were extracted by ethyl acetate, but ZrOCl₂·8H₂O remained unsolved. After filtration and drying, the ZrOCl₂·8H₂O was regenerated to use in the next reaction cycle. The decrease in the reaction yield compared to the number of reaction cycles is shown in Fig. 4.

Fig. 4 Decrease in the reaction yield compared to the number of reaction cycles (at optimized reaction conditions).

Conclusion

In conclusion, we have introduced ZrOCl₂·8H₂O as an efficient, cheap and recyclable catalyst under solvent-free conditions for the one-pot multi-component reaction synthesis of hexahydroquinoline derivatives. Optimization of the reaction conditions was studied using central composite design (CCD). It was shown that by increasing the amount of catalyst, the reaction yield increases, but as temperature increases, a maximum point can be seen in the reaction yield. This behavior can be related to the dehydration/decomposition of ZrOCl₂·8H₂O at temperatures above 85 °C. The quadratic model was best fitted (coefficient of determination = 0.90) to the experimental data. Predicted response values using the obtained model were in good agreement with the experimental results. The promising points for the presented protocol were efficiency, high yields, short reaction times, a cleaner reaction profile and simplicity.

Acknowledgements

The authors gratefully acknowledge partial support of this work by the Research Affairs Office of Bu-Ali Sina University (Grant number 32-1716 entitled development of chemical methods, reagents and molecules), Center of Excellence in Development of Chemical Method (CEDCM), Hamedan, I. R. Iran.

Notes and references

1. I. Ugi, A. Dömling and W. Hörl, Endeavour, 1994, 18, 115–122.
2. L. F. Tietze, Chem. Rev., 1996, 96, 115–136.
3. A. Hantzsch, Ber. Dtsch. Chem. Ges., 1881, 14, 1637–1638.
4. D. M. Stout and A. I. Meyers, Chem. Rev., 1982, 82, 223–243.
5. N. Edraki, A. R. Mehdipour, M. Khoshneviszadeh and R. Miri, Drug Discovery Today, 2009, 14, 1058–1066.
6. T. J. Cleophas and R. van Marum, Am. J. Cardiol., 2001, 87, 487–490.
7. U. Eisner and J. Kuthan, Chem. Rev., 1972, 72, 1–42.
8. D. Schade, M. Lanier, E. Willems, K. Okolotowicz, P. Bushway, C. Wahlquist, C. Gilley, M. Mercola and J. R. Cashman, J. Med. Chem., 2012, 55, 9946–9957.
9. C. Safak and R. Simsek, Mini-Rev. Med. Chem., 2006, 6, 747–755.
10. T. Godfraind, R. Miller and M. Wibo, Pharmacol. Rev., 1986, 38, 321–416.
11. R. Mannhold, B. Jablonka, W. Voigt, K. Schönafinger and E. Schraven, Eur. J. Med. Chem., 1992, 27, 229–235.
12. G. Sabitha, G. S. K. K. Reddy, C. S. Reddy and J. S. Yadav, Tetrahedron Lett., 2003, 44, 4129–4131.
13. M. Maheswara, V. Siddaiah, Y. K. Rao, Y.-M. Tzeng and C. Sridhar, J. Mol. Catal. A: Chem., 2006, 260, 179–180.
14. M. Adharvana Chari and K. Syamasundar, Catal. Commun., 2005, 6, 624–626.
15. E. Perozo-Rondón, V. Calvino-Casilda, R. M. Martín-Aranda, B. Casal, C. J. Durán-Valle and M. L. Rojas-Cervantes, Appl. Surf. Sci., 2006, 252, 6080–6083.
16. L.-M. Wang, J. Sheng, L. Zhang, J.-W. Han, Z.-Y. Fan, H. Tian and C.-T. Qian, Tetrahedron, 2005, 61, 1539–1543.
17. J.-X. Wang, G.-L. Shen and D.-Q. Shi, J. Heterocycl. Chem., 2011, 48, 1145–1148.
18. H. Firouzabadi and M. Jafarpour, J. Iran. Chem. Soc., 2008, 5, 159–183.
19. Z. Zhan-Hui and L. Tong-Shuang, Curr. Org. Chem., 2009, 13, 1–30.
20. M. Ali Zolfigol, A. Khazaei, N. Sarmasti, J. Yousefi Seyf, V. Khakyzadeh and A. R. Moosavi-Zare, J. Mol. Catal. A: Chem., 2014, 393, 142–149.
21. I. C. O. H. (ICH), Guidance for Industry. Q3C Impurities: Residual Solvents, http://www.ich.org/products/guidelines/quality/article/quality-guidelines.html.
22. N. Gorodylova, P. Šulcová, M. Bosacka and E. Filipek, J. Therm. Anal. Calorim., 2014, 118, 1095–1100.
23. M. Kikuchi, Science reports of the Research Institutes. Ser. A, Physics, chemistry and metallurgy, Tohoku University, 1979, vol. 27, p. 81.
24. T. C. W. Mak, Can. J. Chem., 1968, 46, 3491–3497.
25. J. Scholz, K. Scholz and A. J. McQuillan, J. Phys. Chem. A, 2010, 114, 7733–7741.
26. D. A. Powers and H. B. Gray, Inorg. Chem., 1973, 12, 2721–2726.
27. M. A. Zolfigol, A. Khazaei, N. Sarmasti, J. Y. Seyf, V. Khakyzadeh and A. R. Moosavi-Zare, J. Mol. Catal. A: Chem., 2014, 393, 142–149.
28. A. Khazaei, A. R. Moosavi-Zare, H. Afshar-Hezarkhani and V. Khakyzadeh, RSC Adv., 2014, 4, 32142–32147.
29. M. Maheswara, V. Siddaiah, G. L. V. Damu and C. V. Rao, ARKIVOC, 2006, ii, 201–206.
30. G. Sabitha, G. S. Kiran Kumar Reddy, C. S. Reddy and J. S. Yadav, Tetrahedron Lett., 2003, 44, 4129–4131.
31. S.-J. Ji, Z.-Q. Jiang, J. Lu and T.-P. Loh, Synlett, 2004, 5, 831–835.
32. S. B. Sapkal, K. F. Shelke, B. B. Shingate and M. S. Shingare, Tetrahedron Lett., 2009, 50, 1754–1756.
33. S.-J. Song, Z.-X. Shan and Y. Jin, Synth. Commun., 2010, 40, 3067–3077.
34. M. Tajbakhsh, H. Alinezhad, M. Norouzi, S. Baghery and M. Akbari, J. Mol. Liq., 2013, 177, 44–48.
35. S. Ko, M. N. V. Sastry, C. Lin and C.-F. Yao, Tetrahedron Lett., 2005, 46, 5771–5774.
36. M. R. Poor Heravi, S. Mehranfar and N. Shabani, C. R. Chim., 2013 DOI:10.1016/j.crci.2012.11.017.
37. A. Khazaei, M. A. Zolfigol, A. R. Moosavi Zare, J. Afsar, A. Zare, V. Khakyzadeh and M. H. Beyzavi, Chin. J. Catal., 2013, 34, 1936–1944.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra16102a

---