Introduction

New materials are required to be developed for the modern science and technology. These new materials are used for different applications such as drug delivery, medical diagnosis, reinforced composites, semiconductors, electronics, optics, sensors, sorbents, CO2 capture, heterogeneous catalysis, etc. In this manner, nanomaterials can play a vital role1,2,3,4,5,6,7,8,9. One of the emerging fields for the preparation and fabrication of new nanomaterials is dendrimer chemistry which has been recently expanded as two- or three-dimensional covalent organic frameworks (COFs). These strategies afford multifunctional materials which demonstrate synergistic effects and hence, higher performance and efficacy as well as newer and more specific properties than previous counterparts1,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24. In addition, dendrimers can encapsulate and consequently, stabilize metallic catalytic active nanoparticles25,26,27,28. Furthermore, the properties of new materials can be modified and improved by their immobilization onto the surface of magnetic nanoparticles (MNPs), especially in the case of heterogeneous catalysis26,29,30,31,32,33,34,35,36. These improvements include better separation using an external magnetic field34,35,37,38,39,40, enhancement of the reaction rates by MNPs via local heating through induction and increasing the surface area as well as synergistic effects in conjunction with other catalytic species or centers due to the catalytic performance of magnetic materials, including Fe, Ni, Co or Ce-based ones41,42,43. Hence, active catalytic species or centres supported onto the surface of MNPs have received much attention in the field of heterogeneous catalysis for promoting organic reactions in recent years26,44,45,46.

As a particular type of magnetic nanoparticles, superparamagnetic iron oxide nanoparticles (SPIONs) are more widely available than other MNPs due to advantages such as biologically well-accepted constituents, established size-selective preparation, diminished agglomeration, ease of preparation, and lower cost26,45,47,48,49,50,51,52,53,54,55,56,57. On the other hand, heterogenization of the active sites of usual dendritic catalysis has been pursued by either attaching the catalyst covalently within the dendrimer core or at the branch termini as well as through supramolecular interactions such as metal–ligand, hydrogen bonding, aromatic π–π stacking, hydrophobic and van der Waals forces22,26,51,58,59,60,61. Therefore, design and preparation of new magnetic dendritic catalytic systems by appropriate application of dendron segments which can be expanded to 2D or 3D covalent organic frameworks (COFs) is still in high demand.

In recent years, thermally stable heteroaromatic 1,3,5-triazinane-2,4,6‐trione (isocyanurate) moiety has received significant attention in polymer and material chemistry due to its numerous industrial applications, particularly in the field of low toxic drug-delivery agents, tensioactive building blocks and nonlinear optical properties, foams, surface coatings, films, paints, fibers, selective anion receptors and preparation of periodic mesoporous organosilica1,19,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77. On the other hand, boric acid and its derivatives have achieved specific attention, as appropriate catalysts, in organic synthesis due to their advantages including high solubility in water, easy handling, low prices, and environmentally friendly and commercial availability78,79,80,81,82,83,84,85,86,87. In an attempt to indicate how applying SPIONs would affect the dendrimer bearing tridentate and thermally stable isocyanurate moiety as well as boric acid catalytic activity, this study reports the use of multifunctional dendritic nanocatalyst containing boric acid and 1,3,5-tris(2-hydroxyethyl)isocyanurate covalently attached to core–shell silica-coated SPIONs (Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2, 1), as a novel and efficient supramolecular heterogeneous catalyst, in the one-pot synthesis of polyhydroacridines (PHAs, 5) and polyhydroquinolines (PHQs, 7) through multicomponent reaction (MCR) strategy (Scheme 1).

Scheme 1
scheme 1

Schematic representation of the Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2 catalyst (1) and its catalytic activity in the one-pot synthesis of polyhydroacridines (5) and polyhydroquinolines (7) through multicomponent reaction (MCR) strategy (Drawn using the ChemDraw Ultra 12.0 software developed by PerkinElmer).

MCRs are one-pot reactions that involve more than two substrates demonstrating convergence as well as very high atom efficiency and bond-forming-index (BFI)88,89,90. Thus, MCRs are usually a good alternative for the sequential multistep synthesis, especially for useful heterocyclic scaffolds such as Hantzsch esters including 1,4-dihydropyridines (DHPs), PHQs and PHAs in organic synthesis and medicinal chemistry91,92,93,94,95,96. Generally known as one of the main groups of nitrogen heterocycles, polyhydroquinolines (PHQs) and polyhydroacridines (PHAs) have become considerably interesting due to their significant therapeutic and pharmacological properties97,98,99,100. Indeed, they are used as antimalaria, calcium β-blocker, antioxidant, antimicrobial, antifungal, vasodilator, anticancer, bronchodilator, antiatherosclerotic, geroprotective, hepatoprotective and antidiabetic agents as well as in the production of laser colors, radical reservoirs and safe hydrogen transfer agents4,101,102,103,104,105,106,107,108,109,110,111. First introduced by Arthur Hantzsch in 1882, Hantzsch reaction is an MCR that contains the combination of a β-dicarbonyl compound, an aldehyde and a source consisting of ammonia (usually NH4OAc)112. However, catalytic systems are required to accelerate this multicomponent reaction. Here are some recent reported catalysts in this area: Mn@PMO-IL103, vanadium ion doped titania nanoparticles113, Lewis acidic mesoporous material (TUD-1) containing Fe114, magnetite nanoparticle-supported ceria41, silica-coated magnetic nanoparticles with tags of ionic liquid115, Boehmite silica sulfuric acid (Boehmite-SSA) 116, PMO-ICSPrSO3H117, Fe3O4@B-MCM-41118, PS/PTSA119, PdRuNi@GO13, 1,3,5-tris(2-hydroxyethyl) isocyanurate covalently functionalized MCM-41120, alginic acid121,122 and glycine nitrate (GlyNO3) ionic liquid123.

Results and discussion

Characterization of dendritic nanocatalyst containing boric acid and 1,3,5-tris(2-hydroxyethyl)isocyanurate covalently attached to core–shell silica-coated magnetite (Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2, 1)

At first, the boric-acid-functionalized-1,3,5-tris(2-hydroxyethyl)isocyanurate attached to the silica-coated SPIONs (Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2, 1) was characterized using different spectroscopic or analytical methods. As it has been shown in FT-IR spectrum (Fig. 1), the absorption bands at around 632 and 572 cm−1 are related to the Fe–O bond vibrations. On the other hand, absorption band of Si–O–Si asymmetric stretching vibrations are apparent at around 1076 cm−1. Furthermore, the observed signals at 954, 802 and 459 cm–1 are assigned to the symmetric stretching and bending vibrations of Si–O–Si bond43,57,124. Also, the absorption band of C=O bond vibrations of the isocyanurate moiety appeared at around 1637 cm−177,120,125. Furthermore, the signals in range of 1350–1000 cm−1 belong to the C–N bonds vibrations. On the other hand, the absorption band of B–O vibrations appeared at 1510 cm−1. Furthermore, there is an absorption signal at around 1191 cm−1 which is related to B–O–H bond vibrations. Also, the signal at 563 cm−1 is assigned to O–B–O bond vibrations. It is generally accepted that the broad band centred at 3400 cm−1 is ascribed to the stretching vibrations of O–H bonds126,127. All of these data demonstrate that the catalyst 1 has been successfully prepared.

Figure 1
figure 1

The FTIR spectra of Fe3O4, Fe3O4@SiO2, Fe3O4@SiO2@PTS, Fe3O4@SiO2@PTS-THEIC, Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2 (1, from top to down, reproduced using the Microsoft Excel 2016).

Energy dispersive spectroscopy (EDX) spectrum of Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2 (1) proved that the magnetic catalyst functionalized with dendrons containing 1,3,5-tris(2-hydroxyethyl)isocyanurate and boric acid has been functionalized properly due to the presence of Fe, Si, O, C, N and B elements. The percentages of elements were measured to be B (1.96), C (6.99), N (2.50), O (63.58), Si (12.33) and Fe (12.65), respectively. It can be deduced from the absence of Cl and Br elements that terminal chloride groups of the 3-chloropropyl trimethoxysilane (3-APTS) linker as well as terminal bromide groups of the 1,3-dibromopropane linker have been completely replaced by covalent bonding (Fig. 2).

Figure 2
figure 2

Energy dispersive spectroscopy (EDX) analysis of the magnetic Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2 catalyst (1).

The X-ray diffraction (XRD) pattern of Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2 (1) exhibited the phase structure and crystallization of the magnetic nanomaterials (Fig. 3). The main peaks were observed at 2θ: 27.9°, 32.5°, 33.8°, 55.6°, 56.4°, 62.3°. By comparing the XRD pattern of the prepared nanocatalyst (1) with the reference card numbers in the X'pert software, the crystal network of Fe3O4, SiO2 and B(OH)3 correspond with 072–2303, 082–1572 and 030–0199 card numbers, respectively.

Figure 3
figure 3

X-ray diffraction (XRD) pattern of the magnetic dendritic Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2 catalyst (1, the individual reference card numbers of the catalyst 1 components were collected from the X'pert HighScore Plus version 2.1 software developed by the PANalytical B.V.).

The textural properties of the magnetic dendritic Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2 catalyst (1) was investigated by nitrogen adsorption–desorption isotherms (Fig. 4). The BET isotherm of the prepared catalyst corresponds with the BET standard type II adsorption isotherm. The surface area (BET), pore size and pore volume of the catalyst were calculated 55.8 m2/g, 13.9 nm, 0.19 cm3/g, respectively.

Figure 4
figure 4

Nitrogen adsorption–desorption isotherm (BET) of the magnetic Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2 catalyst (1).

Thermal gravimetric analysis (TGA) and differential thermal analysis (DTA) measurements were carried out under air atmosphere by heating the sample at the rate of 10 °C min−1 up to 800 °C (Fig. 5). The first weight loss under 100 °C is related to the removal of water and organic solvents which have remained in the dendritic catalyst through its preparation processes. On the other hand, the second weight loss about 150 °C can be assigned to the dehydration of boric acid moieties and their condensation. Furthermore, two distinct weight losses about 460 and 510 °C are attributed respectively to the decomposition of aliphatic linkers and 1,3,5-tris(2-hydroxyethyl) isocyanurate moieties in the structure of the dendritic Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2 catalyst (1) according to the data obtained by DTA (Fig. 5b).

Figure 5
figure 5

(a) Thermal gravimetric analysis (TGA) and (b) differential thermal analysis (DTA) curves of the magnetic dendritic Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2 catalyst (1).

Vibrating sample magnetometry (VSM) technique was used for measuring the magnetic properties of catalyst (1) at room temperature (Fig. 6). The saturation value of magnetization of Fe3O4 and Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2 was measured to be 47.9 and 35.2 emu/g, respectively. Indeed, the reduction of saturation magnetization of Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2 shows that the dendritic catalyst has been formed. However, the observed saturation magnetization of catalyst (1) is enough and hence, it can be easily separated by an external magnetic field.

Figure 6
figure 6

VSM analysis of the magnetic dendritic Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2 catalyst (1, reproduced using the Microsoft Excel 2016).

To determine the size and morphology of the dendritic Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2 catalyst (1), field emission scanning electron microscopy (FESEM) technique was used (Fig. 7). Interestingly, dendrons containing 3-propyl triethoxysilane (3-PTS), 1,3,5-tris(2-hydroxyethyl)isocyanurate and boric acid moieties are apparent (Fig. 7a–c). This may arise from the combination of both aromatic isocyanurate π-π stacking and boron-oxygen ligand interactions to afford supramolecular arrays of dendrons1,19,128. Furthermore, the obtained images shown in Fig. 7c illustrate that the structure of catalyst was made up of particles smaller than 46 nm.

Figure 7
figure 7

FESEM images of Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2 magnetically recoverable catalyst (1).

Investigation of the catalytic activity of dendritic Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2 (1) for the synthesis of Hantzsch esters

After characterization of the dendritic Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2 catalyst (1), the Hantzsch reaction for the synthesis of polyhydroacridine and polyhydroquinoline derivatives was chosen to examine the catalytic activity of Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2 (1). For this purpose, the condensation of 4-chlorobenzaldehyde (2a, 1 mmol), dimedone (3), NH4OAc (4, 1 mmol) and/or ethyl acetoacetate (6, 1 mmol) were selected as the model reactions, for the synthesis of polyhydroacridine 5a and polyhydroquinoline 7a, respectively. The reactions were optimized considering different parameters such as the amount of catalyst loading, solvents and temperature. The results are reported in Table 1. Indeed, the reaction yield for the desired products 9-(4-chlorophenyl)-3,3,6,6-tetramethyl-3,4,6,7,9,10-hexahydroacridine-1,8(2H,5H)-dione (5a) or ethyl 4-(4-chlorophenyl)-2,7,7-trimethyl-5-oxo-1,4,5,6,7,8-hexahydroquinoline-3-carboxylate (7a) were trace in the absence of any catalyst in EtOH at room temperature (entry 1). However, low yields of the desired products 5a and 7a were obtained under reflux conditions (entry 2) after long times. Interestingly, the yields were improved significantly in the presence of dendritic Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2 catalyst (1, entries 3–5). Further optimization of the reaction conditions illustrated that EtOH is the best solvent to promote the reaction with high efficiency for the synthesis of the desired products 5a or 7a (entries 6–12). The results of optimizing of the model reactions demonstrated that the optimal conditions for the reaction are 10 mg catalyst 1 loading in EtOH under reflux conditions. On the other hand, both boric acid and Fe3O4@SiO2@PTS-THEIC, as the components of the catalyst 1, afforded moderate yields of the desired products 5a and 7a at same catalyst loading under optimized conditions (entries 13 and 14). Finally, hot filtration test (the Sheldon test) was performed to prove the heterogeneous nature of the catalyst 1. During this test, the solid catalyst 1 was removed from the mixture of model reaction for producing 7a by filtration after 10 min using an external magnet. Then, the obtained mixture was heated again for 10 min. The result showed that after removal of the magnetic catalyst 1, the model reaction did not proceed significantly. Indeed, only 48% of the desired product 7a was isolated after 1 h (Fig. 8).

Table 1 Optimization of the reaction of 4-chlorobenzaldeyde (2a), dimedone (3), NH4OAc (4) and/or ethyl acetoacetate (6) under different conditions (The chemical structures were drawn using ChemDraw Ultra 12.0 software developed by PerkinElmer)a.
figure a
Figure 8
figure 8

Hot filtration test for the synthesis of ethyl 4-(4-chlorophenyl)-2,7,7-trimethyl-5-oxo-1,4,5,6,7,8-hexahydroquinoline-3-carboxylate (7a) under optimized conditions (reproduced using the Microsoft Excel 2016).

After finding the optimal conditions, the catalytic activity of Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2 nanocatalyst (1) was further expanded to several other aromatic or heterocyclic aldehydes for the synthesis of other derivatives of PHAs 5a–o and PHQs 7a–u. As it is shown in Tables 2 and 3, the isolated yields of the desired products 6 or 8 were good to excellent in all studied cases under the optimized condition of reaction. In most cases, the products were obtained in similar periods of time and yields compared to the model reaction. Indeed, aldehydes including aromatic carbocyclic or heterocyclic substrates well survived under optimized conditions without formation of any by-products. It is noteworthy that aldehydes bearing electron-withdrawing groups or six-membered heterocycles almost reacted faster than substrates having electron-donating groups or five-membered heterocycles. This trend of reactivity was observed in both symmetric and asymmetric Hantzsch reaction to afford PHAs 5a–o or PHQs 7a–u derivatives, respectively. Furthermore, the α,β-unsaturated cinnamaldehyde (2q) or aliphatic butyraldehyde (2r) reacted in longer reaction times and afforded lower yields. These may be due to resonance and electron-releasing of the double bond and alkyl groups, respectively. All of these findings, led us to purpose a plausible mechanism depicted in Scheme 2.

Table 2 Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2-catalyzed one-pot synthesis of polyhydroacridines 5a–o from different aldehydes (2a–o), dimedone (3) and NH4OAc (4) under the optimized conditions (The chemical structures were drawn using ChemDraw Ultra 12.0 software developed by PerkinElmer)a.
figure b
Table 3 Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2-catalyzed one-pot synthesis of polyhydroquinolines 7a–u from different aldehydes (2a–u), dimedone (3), NH4OAc (4) and ethyl acetoacetate (5) under the optimized conditions (The chemical structures were drawn using ChemDraw Ultra 12.0 software developed by PerkinElmer)a.
figure r
Scheme 2
scheme 2

Plausible mechanism for the one-pot synthesis of polyhydroacridines 5 and polyhydroquinolines 7 catalyzed by the agnetically recoverable Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2 catalyst (1, Drawn using the ChemDraw Ultra 12.0 software developed by PerkinElmer).

An important distinguishing feature of this magnetic dendritic nanocatalyst (1) beside easy separation from the reaction mixture is its recyclability. After the reaction was completed, the catalyst was separated and washed by acetone and hexane, respectively. Then, it was dried and reused in the model reactions for the next runs. The obtained results have been summarized in Fig. 9. These results show that this catalyst can be recovered and reused at least for five times in further runs under optimized conditions without a notable loss of its activity. Furthermore, comparison of the FTIR spectra of both fresh dendritic Fe3O4@SiO2@CPTS-THEIC-(CH2)3OB(OH)2 nanocatalyst (1) and the recycled sample after six consecutive runs for the synthesis of 5a demonstrated that their structures are almost similar (Fig. 10).

Figure 9
figure 9

Recyclability of the dendritic Fe3O4@SiO2@CPTS-THEIC-(CH2)3OB(OH)2 nanocatalyst (1) for the synthesis of 5a and 7a (Drawn using the Microsoft Excel 2016).

Figure 10
figure 10

FTIR spectra of the fresh Fe3O4@SiO2@CPTS-THEIC-(CH2)3OB(OH)2 nanocatalyst (1) and the recycled sample after six consecutive runs for the synthesis of 5a (reproduced using the Microsoft Excel 2016).

Table 4 contains some of the formerly reported methods and representing their catalytic activity for the synthesis of polyhydroacridines and polyhydroquinolines to compare them with the dendritic Fe3O4@SiO2@CPTS-THEIC-(CH2)3OB(OH)2. These data clearly demonstrate that the nanocatalyst 1 is more active than other previously reported catalytic systems in terms of catalyst loading, product yield, required reaction time and avoiding the toxic solvents.

Table 4 Comparison of the synthesis of compounds 5a and 7a using the reported methods versus the present method.

Experimental section

General information

All chemicals and reagents were provided by Merck or Aldrich chemical companies and used as received without any further purification, except for benzaldehyde which was used as a fresh distilled sample. FTIR spectra were recorded using KBr pellets on a Shimadzu FT IR-8400S spectrometer. Energy dispersive spectroscopy (EDS) was recorded on a SAMx instrument. The X-ray powder diffraction (XRD) data were collected on an X'Pert MPD Philips diffractometer with Cu radiation source (λ = 1.54050 Å) at 40 kV voltage and 40 mA current. Field emission scanning electron microscopy (FESEM) images were obtained using a MIRA3 instrument of TESCAN Company, Czech Republic. Thermal gravimetric analysis (TGA) and differential thermal analysis (DTA) were performed by means of a Bahr company STA 504 instrument. The BET specific surface area of the catalyst 1 was obtained using an equipment ASAP 2020 Micromeritics. Magnetic susceptibility measurements were taken out by using a Lakeshore VSM, 7410 series. Melting points were determined using an Electrothermal 9100 apparatus and are uncorrected. 1H NMR (500 MHz) spectra were obtained using a Bruker DRX-500 AVANCE spectrometer in CDCl3 at ambient temperature. Analytical TLC was carried out using Merck 0.2 mm silica gel 60 F-254 Al-plates and n-hexane: EtOAc, (3:1, v/v %) as eluent. All products are known and their structures were established by comparing the physical constants as well as FTIR and NMR spectroscopic data with authentic samples120,122,141.

Preparation of Fe 3 O 4 @SiO 2 nanoparticles modified by (3-chloropropyl) trimethoxysilane (Fe 3 O 4 @SiO 2 @CPTS)

The Fe3O4@SiO2@CPTS materials were prepared according to the reported methods in literature with a slight modification56.

Preparation of the dendritic Fe 3 O 4 @SiO 2 @CPTS@THEIC nanomaterials

Fe3O4@SiO2@CPTS (1 g) was dispersed in toluene (30 ml) and KI (1.66 g) was added to the obtained mixture with the mechanical stirring at 80 °C for 1 h. Then, K2CO3 (1.38 g) and tris-(2-hydroxyethyl)-1,3,5-triazinane-2,4,6-trione (1 g) were added to the mixture and it was heated under reflux conditions for 8 h. The obtained solid was filtered off and washed with EtOH (5 ml) and then dried in an oven for 2 h.

Preparation of the dendritic Fe 3 O 4 @SiO 2 @PTS-THEIC-(CH 2 ) 3 OB(OH) 2 nanocatalyst (1)

A mixture of Fe3O4@SiO2@CPTS@THEIC (1 g) and 1,3-dibromopropane (d = 1.98 g.cm−3, 2 ml) was added to toluene (15 ml) and heated at 40 °C for 12 h. The obtained solid was filtered off, washed with toluene (5 ml) and then dried in a vacuum oven at 60 °C for 2 h. The as-prepared solid and H3BO3 (1 g) were mixed in EtOH (30 ml) and the obtained mixture was stirred at room temperature for 18 h. After completion of the process, the obtained brown solid was filtered off and washed with EtOH (5 ml) on a Buchner funnel and then kept in a vacuum oven at 60 °C for 12 h. The complete procedure for the preparation of catalyst 1 has been represented in Scheme 3.

Scheme 3
scheme 3

Schematic preparation of the dendritic Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2 catalyst (1, Drawn using the ChemDraw Ultra 12.0 software developed by PerkinElmer).

General procedure for the synthesis of 1,8-dioxoacridindione derivatives 5a–o catalyzed by magnetic dendritic Fe 3 O 4 @SiO 2 @PTS-THEIC-(CH 2 ) 3 OB(OH) 2 catalyst (1)

In a 5 mL round-bottomed flask, a mixture of aldehyde (2a–o, 1 mmol), dimedone (3, 2 mmol, 0.28 g), NH4OAc (4, 1.5 mmol, 0.11 g) and Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2 (1, 0.01 g) were added to EtOH 96% (2 mL). The obtained mixture was stirred under reflux conditions for the times indicated in Table 2. The progress of the reactions was monitored by TLC experiment (eluent; n-hexane: EtOAc, 3:1, v/v %). After completion of the reaction, EtOH (3 mL) was added to the mixture and it was heated to dissolve all organic compounds. Then, the catalyst 1 was easily separated by an external magnet and the solution was filtered. The filtrate was kept at room temperature and the crystals were collected by filtration to afford 1,8-dioxoacridindione derivatives 5a–o in high purity.

General procedure for the synthesis of polyhydroquinoline derivatives 7a–u catalyzed by magnetic dendritic Fe 3 O 4 @SiO 2 @PTS-THEIC-(CH 2 ) 3 OB(OH) 2 catalyst (1)

In a 5 mL round-bottomed flask, a mixture of aldehyde (2a–u, 1 mmol), dimedone (3, 1 mmol, 0.14 g), NH4OAc (4, 1.5 mmol, 0.11 g), ethyl acetoacetate (5, 1 mmol, 0.13 g) and Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2 (1, 0.01 g) were added to EtOH 96% (2 ml). The obtained mixture was stirred under reflux conditions for times indicated in Table 3. The progress of the reactions was monitored by TLC experiment (eluent; n-hexane: EtOAc, 3:1, v/v %). After completion of the reaction, EtOH (3 mL) was added to the mixture and it was heated to dissolve all organic compounds. Then, the catalyst 1 was easily separated by an external magnet and the solution was filtered. The filtrate was kept at room temperature and the crystals were collected by filtration to afford polyhydroquinoline derivatives 7a–u in high purity.

Selected spectral data

9-(4-Chlorophenyl)-3,3,6,6-tetramethyl-3,4,6,7,9,10-hexahydro-1,8(2H,5H)-acridinedione (5a)

Pale yellow solid; m.p. = 310–312 °C; FT-IR (KBr, cm−1): 3282, 3176, 3060, 2954, 2875,1650, 1608, 1492, 1365, 1220, 1147, 1089, 1014, 840, 761, 597, 526; 1H NMR (500 MHz, CDCl3): δ (ppm): 0.98 (s, 6H, 2CH3), 1.10 (s, 6H, 2CH3), 2.19–2.37 (8H, m, 4CH2), 5.06 (s, 1H, CH), 7.17 (d, 2H, Ar–H), 7.28 (d, 2H, Ar–H), 6.97 (s, 1H, NH).

3,3,6,6-Tetramethyl-9-(pyridin-2-yl)-3,4,6,7,9,10-hexahydroacridine-1,8(2H,5H)-dione (5i):

Pale yellow solid; m.p. = 284–286 °C; FT-IR (KBr, cm−1): 3604, 3519, 3440, 3284, 2875, 1637, 1600, 1477, 1365, 1218, 1139, 995, 744, 563; 1H NMR (500 MHz, CDCl3): δ (ppm): 0.98 (s, 6H, 2CH3), 1.07 (s, 6H, 2CH3), 2.12–2.46 (8H, m, 4CH2), 5.22 (s, 1H, CH), 7.51–7.58 (t, 3H, Ar–H), 8.41 (d, 1H, Ar–H), 6.97 (s, 1H, NH).

Ethyl 4-(4-methoxyphenyl)-2,7,7-trimethyl-5-oxo-1,4,5,6,7,8-hexahydroquinoline-3-carboxylate (7l)

Pale yellow solid; mp.: 255–260 °C; FT-IR (KBr, cm−1): 3278, 3203, 3076, 2956, 1699, 1604, 1496, 1379, 1276, 1218, 1070, 1031, 842, 765, 536; 1H NMR (500 MHz, CDCl3): δ (ppm): 0.92 (s, 3H, CH3), 1.04 (s, 3H, CH3), 1.20 (t, 3H, J = 7.2 Hz, CH3 (OEt)), 2.11–2.28 (m, 4H, CH2), 2.33 (s, 3H, CH3), 3.71 (s, 3H, OCH3), 4.03–4.07 (q, 2H, J = 7.2 Hz, CH2 (OEt)), 4.98 (s, 1H, CH benzylic), 6.43 (br s, 1H, NH), 6.71–6.73 (d, 2H, J = 8.2 Hz, Ar–H), 7.21 (d, 2H, J = 8.2 Hz, Ar–H).

Ethyl 2,7,7-trimethyl-4-(3-nitrophenyl)-5-oxo-1,4,5,6,7,8-hexahydroquinoline-3-carboxylate (7e)

Pale yellow solid; m.p. = 180–184 °C; FT-IR (KBr, cm−1): 3276, 3193, 2964, 1703, 1604, 1490, 1379, 1278, 1215, 1143, 1070, 1022, 829, 754, 690, 507; 1H NMR (500 MHz, CDCl3): δ (ppm): 0.93 (s, 3H, CH3), 1.09 (s, 3H, CH3), 1.19 (t, 3H, J = 7.2 Hz, CH3 (OEt)), 2.13–2.40 ( 7H, s CH3, m 2CH2), 4.03–4.07 (q, 2H, J = 7.2 Hz, CH2 (OEt)), 5.15 (s, 1H, CHbenzylic), 5.98 (s, 1H, NH), 7.35–8.10 (m, 2H, Ar–H) (Supplementary Information 1).

Conclusions

In conclusion, the multifunctional dendritic nanocatalyst containing boric acid and 1,3,5-tris(2-hydroxyethyl)isocyanurate covalently attached to core–shell silica-coated magnetite (Fe3O4@SiO2@PTS-THEIC-(CH2)3OB(OH)2) was prepared and properly characterized for the first time. It was found that the combination of both aromatic π–π stacking and boron–oxygen ligand interactions affords supramolecular arrays of dendrons. The use of boric acid makes this dendritic catalyst a green choice from corrosion, recyclability and cost points of view. The magnetic dendritic catalyst was used, as a mild and recyclable catalyst, for the one-pot efficient synthesis of polyhydroacridines and polyhydroquinolines through MCR strategy in EtOH as a green solvent. Indeed, very low catalyst loading, short reaction times, mild reaction conditions, high to excellent yields, reusability of the catalyst, ease of separation by an external magnetic field, and the use of nontoxic materials for the preparation of the catalyst are among other advantages of this protocol. Further exploring of this magnetic dendritic magnetic catalyst for other organic transformations is underway in our research lab and would be presented in due course.