Inorganic–organic hybrid sorbent for aromatic desulfurization of hydrocarbons: regenerative adsorption based on a charge-transfer complex

Abstract

A new series of inorganic–organic hybrid materials as sorbents for the adsorption of dibenzothiophene (DBT) and 4,6-dimethyldibenzothiophene (4,6-DMDBT) have been prepared by functionalizing mesoporous silica with 9,10-anthraquinone derivatives. A selected sorbent was fully characterized via ¹³C and ²⁹Si cross-polarization magic-angle solid-state nuclear magnetic resonance (CP-MAS NMR), FTIR spectroscopy, powder X-ray diffraction (PXRD) and thermal analysis (TGA-DTA). The prepared materials were used to adsorb DBT and 4,6-DMDBT from C₇ hydrocarbons under ambient conditions. The results show that the hybrid material prepared by the functionalization of mesoporous silica with 9,10-dioxo-5-((3-(triethoxysilyl)-propyl)amino)-1,5-dichloro-4,8-dinitroanthraquinone is an efficient and regenerable sorbent for adsorption of DBT and 4,6-DMDBT by π–π interaction. To demonstrate the adsorption of DBT and 4,6-DMDBT through π–π interaction, a charge-transfer complex (CTC) of 1,5-dichloro-4,8-dinitroanthraquinone and 4,6-DMDBT was prepared and characterized by FT-IR, ¹H and ¹³C NMR spectroscopy, MS, TGA-DTA and single-crystal X-ray diffraction. The results show that 1,5-dichloro-4,8-dinitroanthraquinone produces a CTC with 4,6-DMDBT, and when it is grafted on Si-MCM-41, it produces a hybrid material with the capability of adsorbing only aromatic DBT and 4,6-DMDBT rather than anthracene, naphthalene and xylene. In contrast, 1,5-dichloroanthraquinone and 1,5-dicyanoanthraquinone neither produce a CTC nor adsorb DBT, 4,6-DMDBT, anthracene, naphthalene and xylene after they are grafted on Si-MCM-41. This study proves that the hybrid material prepared from mesoporous silica and 9,10-dioxo-5-((3-(triethoxysilyl)-propyl)amino)-1,5-dichloro-4,8-dinitroanthraquinone is an efficient, environmentally benign and regenerable sorbent for the removal of DBT and 4,6-DMDBT from C₇ hydrocarbons.

---

Introduction

The ultra-deep desulfurization of gasoline and diesel fuel continuously thrives due to the environmental impacts of these fuels.^1–6 The majority of global petroleum reserves contain sour crudes, and there is great demand to improve the efficiency of the desulfurization technology.⁷ According to new environmental regulations in several countries, the sulphur level in gasoline and diesel fuel must be reduced to below 10 ppm, which by definition is called “S-free” fuel.⁸ Furthermore, the reduction of CO and NO_x in lean gasoline burning engines equipped with catalytic converters requires ultra-low sulphur fuels to increase the lifetime of the catalyst and efficiency of the engines.⁹ In fuel-cell technology, liquid fuels, such as gasoline or diesel oil, also are important sources for fuel gas (H₂), which requires almost S-free fuels.¹⁰ The present state of art desulfurization technology is based on hydrodesulfurization (HDS), and in this process, which is based on CoMo or NiMo catalysts, organic sulphur compounds are converted to H₂S and hydrocarbons.¹¹ This technology, due to the reaction conditions, requires high pressure and temperature reactors and vessels. Consequently, it requires heavy capital investment, which makes the process very expensive. Furthermore, it is suitable only for removing paraffinic compounds such as thiols, thioethers and disulphides. The HDS process suffers from low reactivity of cyclic and particularly aromatic sulphur compounds.¹² It is well-known that after conventional HDS, a significant amount of benzothiophene (BT), dibenzothiophene (DBT), 4,6-dimethydibenzothiophene (4,6-DMDBT) and similar derivatives along with aromatic hydrocarbons remain in the diesel fraction.^13,14 Therefore, various desulfurization methods, such as extraction using ionic liquids,^15–17 oxidation desulfurization,^18,19 biodesulfurization²⁰ and adsorptive desulfurization, have been developed.^21,22 Adsorptive methods that use various sorbents, including activated carbon, zeolites,^23,24 transition metal supported on alumina,^25–27 modified mesoporous silica,²⁸ adsorption by π-complexation with supported metal ions^29–35 and charge transfer complex (CTC) adsorption, have been introduced in the literature with interesting results.^36–41 However, their efficiency for removing aromatic organosulfurs from diesel fuel is low, and developing an alternative green desulfurization process with better performance and without the need for a solvent and severe reaction conditions, such as pressure and temperature, should be beneficial. It is well-known that electron-rich aromatic π-systems can interact with electron deficient π-systems through π–π stacking, and interestingly such stacking can be disassembled by small activation energies.^42–44 Therefore, it should be possible to use such π–π interactions in reversible adsorption desorption processes. Weak interactions between dibenzothiophene derivatives, i.e. electron-rich π-systems, and π-acceptors to form a CTC offer an opportunity to remove these compounds from fossil fuel after hydrodesulfurization. This π–π interaction has been implemented in the removal of 4,6-dialkyldibenzothiophenes from straight run Arabian Light by immobilizing 4,5-dicyano-2,7-dinitrofluorenone or 2,4,5,7-tetranitro-9-fluorenone on poly(styrene-co-divinylbenzene), and 2,4,5,7-tetranitro-9-fluorenylideneaminooxy propionic acid on silica from n-heptane.^37,41,45 However, in spite of impressive results, the use of 4,5-dicyano-2,7-dinitrofluorenone, 2,4,5,7-tetranitro-9-fluorenone and similar derivatives on an industrial scale is doubtful due to their very high cost. Furthermore, the regeneration of sorbents using toluene as a solvent is inevitable. In this context, the development of new materials for the removal of refractory sulphur compounds from fuel by the same mechanism can transform the commercialization of this process into reality.

In an earlier report, we demonstrated the formation of a CTC between the electron-deficient 1,8-dihydroxy-2,4,5,7-tetranitro-9,10-anthraquinone and electron-rich dibenzothiophene, and 4,6-dimethydibenzothiophene.⁴⁶ In continuation of our interest in the ultra-deep desulfurization of hydrocarbons, as a model for gasoline or diesel fuel, we graft 1,5-dichloroanthraquinone (1a), 1,5-dicyanoanthraquinone (1b) and 1,5-dichloro-4,8-dinitroanthraquinone (1c) as electron-deficient π-systems on MCM-41 mesoporous silica as sorbents for dibenzothiophene (DBT) and 4,6-dimethyldibenzothiophene (4,6-DMDBT). Mesoporous MCM-41 silica, which has a high surface area, large pore volume, ease of preparation and most importantly is commercially available in a sizeable volume, makes a unique support when compared to other mesoporous silicas.

To graft an electron-deficient π-system on mesoporous silica and prepare inorganic–organic hybrid materials, three new 3-aminopropyltriethoxysilane derivatives, 1-chloro-5-((3-(triethoxysilyl)propyl)amino)anthracene-9,10-dione (2a), 9,10-dioxo-5-((3-(triethoxysilyl)propyl)amino)-9,10-dihydroanthracene-1-carbonitrile (2b) and 1,5-dichloro-4-nitro-8-((3-(triethoxysilyl)propyl)amino)anthracene-9,10-dione (2c) in addition to 1-chloro-5-((3-(trimethoxysilyl)propyl)amino)anthracene-9,10-dione (2d) were synthesized by the reaction of 3-aminopropyltriethoxysilane and/or 3-aminopropyltrimethoxysilane with 1a–1c (Scheme 1). All 3-aminopropyltriethoxysilane anthracene-9,10-dione derivatives, 2a–2c, were fully characterized and then grafted on mesoporous Si-MCM-41.

Scheme 1 Synthesis of 2a–d from 1a–c.

Inorganic–organic hybrid materials with electron-deficient π-systems were used as sorbents to study the removal of DBT and 4,6-DMDBT from C₇ hydrocarbons. To demonstrate the π–π interaction in the adsorption of DBT and 4,6-DMDBT in these inorganic–organic hybrid materials, the formation of a CTC between the electron-deficient π-systems, 1a–c, and DBT and 4,6-DMDBT was investigated. The CTC that was formed between 1,5-dichloro-4,8-dinitro-anthraquinone and 4,6-dimethyldibenzothiophene (1,5-dichloro-4,8-dinitro-anthraquinone:4,6-dimethydibenzothiophene) was characterized by single-crystal X-ray diffraction in addition to spectroscopic measurements and thermal analyses.

Experimental

Materials and methods

Tetraethoxysilane, cetyltrimethylammonium bromide, 3-aminopropyltriethoxysilane, 3-aminopropyltrimethoxysilane, 1,5-dichloroanthraquinone, dibenzothiophene, and 4,6-dimethyldibenzothiophene were obtained from Aldrich. Anthracene, naphthalene, xylene and all solvents were purchased from Merck and used without purification. 4,6-Dimethyldibenzothiophene was obtained from Aldrich. All solvents were dried and distilled under nitrogen prior to use according to the standard procedure.⁴⁷ 1,5-Dicyanoanthraquinone (1b) and 1,5-dichloro-4,8-dinitroanthraquinone (1c) were prepared from 1,5-dichloroanthraquinone according to the earlier reports.^48,49 The Si-MCM-41 mesoporous silica was synthesized according to the previously reported procedure,⁵⁰ and its formation was confirmed by low-angle X-ray powder diffraction and FTIR spectroscopy. Elemental analyses were performed with a Thermo Finnigan Flash-1112EA microanalyzer. ¹H and ¹³C {¹H} NMR spectra were obtained at room temperature in CDCl₃ or DMSO-d₆ on a Bruker AVANCE 300 MHz instrument operating at 300.3 and 75.4 MHz, respectively. NMR spectra are referenced to Me₄Si as an external standard. Infrared spectra from 4000–400 cm⁻¹ were acquired on a Shimadzu 470 FT-IR instrument using KBr pellets. Mass spectra (ESI) were obtained on a Finnigan LCQ mass spectrometer. Thermal analyses were carried out on a BAHR STA-503 instrument under ambient conditions at a heating rate of 10 °C min⁻¹ in air. X-ray powder diffraction patterns were obtained on a STOE diffractometer with Cu K_α radiation at 60 keV and 15 mA with a scanning rate of 3° min⁻¹ in the 2θ range from 1° to 10°. ¹³C and ²⁹Si NMR solid-state NMR spectra were obtained on a Bruker Varian 400 VRX spectrometer operating at 100.5 and 108.1 MHz resonance frequency for ¹³C and ²⁹Si, respectively, with a conventional double resonance 5 mm CP/MAS probe. Single-crystal X-ray diffraction data were collected on a Bruker SMART APEX with graphite monochromated Mo K_α radiation using the APEX2 software.⁵¹ Structures were solved by direct and subsequent difference Fourier map and then refined on F² by a full-matrix least-squares procedure using anisotropic displacement parameters methods.⁵² All refinements were performed using the SHELXL97 crystallographic software package.⁵³

Synthesis of anthraquinone derivatives (2a–d)

For the synthesis of anthraquinone derivatives (2a–d), 3-aminopropyltriethoxysilane or 3-aminopropyltrimethoxysilane (10 mmol) was added dropwise to a slurry of anthraquinone (1a–c) (10 mmol) in dry toluene or xylene (50 mL) and then the mixture was refluxed under an N₂ atmosphere. The progress of the reaction was monitored by TLC with ethyl acetate/n-hexane (1 : 3). After 24 h, the solution was allowed to cool and then filtered. The solvent was removed under reduced pressure and the residue was crystallized from n-hexane to afford the final product.

1-Chloro-5-((3-(triethoxysilyl)propyl)amino)anthracene-9,10-dione (2a, AQ-Cl)

Red solid (3.0 g, 65%), mp 98–99 °C. Anal. calcd for C₂₃H₂₈ClNO₅Si: C, 59.79; H, 6.11; N, 3.03; found: C, 59.82; H, 6.05; N, 2.95%. IR (KBr) (v_max/cm⁻¹): 3257, 3446, 3257, 2973, 2929, 2885, 1666, 1630, 1592, 1574, 1510, 1474, 1445, 1405, 1360, 1305, 1261, 1191, 1163, 1105, 1081, 1003, 959, 804, 762, 706, 607, 545, 470, 424. ¹H NMR (300.13 MHz, CDCl₃) δ_H: 0.79 (t, ³J_H–H = 8.0 Hz, 2H, CH₂), 1.25 (t, ³J_H–H = 7.0 Hz, 9H, CH₃) 1.86–1.91 (m, 2H, CH₂), 3.37 (q, ³J_H–H = 7.0 Hz, 2H, CH₂), 3.86 (q, ³J_H–H = 7.0 Hz, 6H, OCH₂), 7.05–7.09 (m, 1H, H–Ar), 7.53–7.56 (m, 2H, H–Ar), 7.62–7.67 (m, 1H, H–Ar); 7.70–7.73 (m, 1H, H–Ar), 8.29–8.32 (m, 1H, H–Ar), 9.68 (broad, 1H, NH). ¹³C {¹H} NMR (75.47 MHz, CDCl₃) δ_C: 7.9, 18.3, 22.7, 45.4, 58.4, 112.1, 115.8, 117.3, 126.1, 129.0, 129.4, 134.0, 134.4, 135.6, 135.8, 137.2, 151.5, 182.6, 183.2. ²⁹Si NMR (59.6 MHz, CDCl₃) δSi: –45.9. MS (ESI) (m/z) (methanol): 370.18 [M–EtOH–EtO]⁺, 463.22 [M + H]⁺, 555.19 [M + 2EtOH + H]⁺.

9,10-Dioxo-5-((3-(triethoxysilyl)propyl)amino)-9,10-dihydroanthracene-carbonitrile (2b, AQ-CN)

Red-brown solid (2.25 g, 49%), mp 155 °C. Anal. calcd for C₂₄H₂₈N₂O₅Si: C, 63.69; H, 6.24; N, 6.19; found: C, 63.73; H, 6.10; N, 6.13%. IR (KBr) (v_max/cm⁻¹): 3428, 3288, 3091, 2974, 2930, 2886, 2223, 1668, 1628, 1592, 1571, 1507, 1464, 1409, 1325, 1290, 1166, 1103, 1078, 956, 815, 852, 765, 705, 608, 545, 471, 425.¹H NMR (300.13 MHz, CDCl₃) δ_H: 0.79 (t, 2H, ³J_H–H = 8.0 Hz, CH₂), 1.25 (t, 9H, ³J_H–H = 7.0 Hz, 3CH₃) 1.68–1.73 (m, 2H, CH₂), 3.37 (q, 2H, ³J_H–H = 7.0 Hz, CH₂), 3.86 (q, 6H, ³J_H–H = 7.0 Hz, 3OCH₂), 7.05–7.09 (m, 1H, H–Ar), 7.53–7.56 (m, 2H, H–Ar), 7.62–7.67 (m, 1H, H–Ar); 7.70–7.73 (m, 1H, H–Ar), 8.29–8.32 (m, 1H, H–Ar). ¹³C {¹H} NMR (75.47 MHz, CDCl₃) δ_C: 7.8, 18.3, 22.7, 45.4, 58.5, 118.9, 120.6, 122.7, 124.3, 126.8, 127.4, 129.0, 130.6, 131.0, 136.6, 137.1, 138.6, 150.3, 180.8 and 180.9. ²⁹Si NMR (59.6 MHz, CDCl₃) δSi: −46.1. MS (ESI) (m/z) (methanol): 927 [2M + Na]⁺.

1,5-Dichloro-4-nitro-8-((3-(ethoxysilyl)propyl)amino)anthracene-9,10-dione (2c, AQ-NO₂)

Violet solid (1.12 g, 20%), mp 102 °C. Anal. calcd for C₂₃H₂₆Cl₂N₂O₇Si: C, 51.02; H, 4.84; N, 5.17; found: C, 51.24; H, 4.82; N, 5.47%. IR (KBr) (v_max/cm⁻¹): 3435, 3071, 2973, 2926, 2887, 1684, 1634, 1597, 1570, 1546, 1502, 1466, 1372, 1339, 1309, 1234, 1196, 1169, 1101, 1076, 957, 815, 767, 657, 557 and 491. ¹H NMR (300.13 MHz, CDCl₃) δ_H: 0.76 (t, ³J_H–H = 7.7 Hz, 2H, CH₂), 1.26 (t, ³J_H–H = 7.0 Hz, 9H, CH₃) 1.82–1.90 (m, 2H, CH₂), 3.38 (q, ³J_H–H = 6.9 Hz, 2H, CH₂), 3.86 (q, ³J_H–H = 7.0 Hz, 6H, OCH₂), 7.06 (d, ³J_H–H = 9 Hz, 1H, H–Ar), 7.50 (d, ³J_H–H = 9 Hz, 1H, H–Ar), 7.69 (d, ³J_H–H = 9 Hz, 1H, H–Ar); 7.81 (d, ³J_H–H = 9 Hz, 1H, H–Ar), 9.58 (broad, 1H, NH). ¹³C {¹H} NMR (75.47 MHz, CDCl₃) δ_C: 7.8, 18.4, 22.7, 45.4, 58.5, 118.9, 120.7, 122.7, 124.3, 126.6, 127.4, 129.0, 130.3, 130.6, 137.1, 138.6, 150.3, 180.8, 180.9. ²⁹Si NMR (59.6 MHz, CDCl₃) δSi: −46.2. MS (ESI) (m/z) (methanol): 449.07 [M–2EtOH]⁺, 540.84 [M + H]⁺, 634.16 [M + 2EtOH + H]⁺, 726.01 [M + 4EtOH + H]⁺, 1104.59 [2M + Na]⁺.

1-Chloro-5-((3-(trimethoxysilyl)propyl)amino)anthracene-9,10-dione (2d)

Red solid (2.80 g, 67%), mp 180 °C. Anal. calcd for C₂₀H₂₂ClNO₅Si: C, 57.20; H, 5.28; N, 3.34; found: C, 57.61; H, 5.57; N, 3.20%. IR (KBr) (v_max/cm⁻¹): 3430, 3268, 2938, 2838, 1668, 1629, 1593, 1571, 1509, 1474, 1406, 1305, 1266, 1193, 1160, 1078, 1027, 824, 787, 757, 703, 587, 547, 460. ¹H NMR (300.13 MHz, CDCl₃) δ_H: 0.80 (t, ³J_H–H = 8.0 Hz, 2H, CH₂), 1.83–1.91 (m, 2H, CH₂), 3.35 (q, ³J_H–H = 8.0 Hz, 2H, CH₂), 3.60 (s, 9H, OCH₃), 7.03–7.06 (m, 1H, H–Ar), 7.53–7.55 (m, 2H, H–Ar), 7.61–7.65 (m, 1H, H–Ar), 7.68–7.71 (m, 1H, H–Ar), 8.26–8.29 (m, 1H, H–Ar), 9.68 (broad, 1H, NH). ¹³C {¹H} NMR (75.47 MHz, CDCl₃) δ_C: 7.3, 22.9, 45.7, 51.0, 112.6, 116.0, 117.7, 126.6, 127.4, 133.7, 134.4, 136.1, 136.7, 137.6, 138.0, 151.1, 183.0, 183.6. ²⁹Si NMR (59.6 MHz, CDCl₃) δSi: −46.7. MS (ESI) (m/z) (methanol): 860.65 [2M + Na]⁺. The suitable crystals of 2d for single-crystal X-ray crystallography were obtained by slow evaporation of heptane solution of 2d in one month. Unit cell parameters: a = 7.145(5) Å, b = 35.461(16) Å, c = 10.506(7) Å, α = 90°, β = 133.00(11)°, γ = 90°; space group P2₁/c, R = 0.18, wR = 0.43.

Preparation of 4,6-dimethyldibenzothiophene:1,5-dichloro-4,8-dinitroanthraquinone charge transfer complex (CTC)

A solution of 4,6-dimethydibenzothiophene (0.18 g, 1 mmol) in DMF (5 mL) was added to a solution of 1,5-dichloro-4,8-dinitroanthraquinone (0.36 g, 1 mmol) in the same solvent (15 mL). After filtration, the solution was kept in a tightly closed vial at −4 °C in a freezer. Dark-red single-crystals were collected from the bottom of the vial after 4 days (yield 80%). Anal. calcd for C₂₈H₁₆Cl₂N₂O₆S: C 58.04, H 2.78, N 4.83; found: C 58.31, H 2.57, N 5.13%. IR (KBr) (v_max/cm⁻¹): 3420, 3087, 3063, 2922, 2852, 1743, 1686, 1571, 1543, 1440, 1398, 1379, 1308, 1242, 1194, 1142, 1050, 850, 822, 768, 727, 647, 589, 566, 471, 405. ¹H NMR (300.13 MHz, DMSO-d₆) δ_H: 2.53 (3H, s, CH₃), 7.35–7.47 (2H, m, H–Ar), 8.15–8.25 (3H, m, H–Ar). ¹³C {¹H} (75.47 MHz, DMSO-d₆) δ_C: 20.5 (CH₃), 120.2, 125.7, 127.7, 129.2, 129.3, 130.5, 132.3, 135.6, 135.9, 138.1, 138.5, 147.3, 177.7. MS (ESI) (m/z) (CH₃OH/DMF): 214.1 (C₁₄H₁₂S + 2H)⁻, 311.4 (C₁₄H₄ClNO₄ + Na + 3H)⁻, 325.4 (C₁₄H₄ClN₂O₄ + Na + 3H)⁻, 392.2 (C₁₄H₄Cl₂N₂O₆ + Na + 2H)⁻. Single-crystal data: a = 10.6317(8) Å, b = 14.5110(10) Å, c = 16.1526(9) Å, α = β = γ = 90°; space group Pcnb, wR1 = 0.061, wR2 = 0.142.

Grafting 3-aminopropyltriethoxysilane anthracene-9,10-dione derivatives on Si-MCM-41

In a typical reaction, 1.0 g of Si-MCM-41 was suspended in 50 mL toluene and the mixture was stirred at room temperature for 1 h, and then 0.10 and/or 0.20 mmol π-accepter compound, 1-chloro-5-((3-(triethoxysilyl)propyl)amino)anthracene-9,10-dione (2a, AQ-C), 9,10-dioxo-5-((3-(triethoxysilyl)propyl)amino)-9,10-dihydroanthracene-carbonitrile (2b, AQ-CN) and/or 1,5-dichloro-4-nitro-8-((3-(ethoxysilyl)propyl)amino)anthracene-9,10-dione (2c, AQ-NO₂) were added and the mixture was refluxed overnight. The solid was removed from the solvent by filtration and washed with toluene and dichloromethane and then dried at room temperature. Grafting of a selected Si-MCM-41 sample was established by ¹³C and ²⁹Si CP-MAS NMR, and FT-IR spectroscopy, small-angle X-ray powder diffraction, and thermal and elemental analyses. The grafted materials are designated as AQ-Cl@Si-MCM-41, AQ-CN@Si-MCM-41 and AQ-NO₂@Si-MCM-41.

Adsorption of DBT and 4,6-DMDBT by various sorbents

For adsorption studies, 100 mg of each of the electron-deficient π-system functionalized Si-MCM-41 (AQ-Cl@Si-MCM-41, AQ-CN@Si-MCM-41 and AQ-NO₂@Si-MCM-41) was suspended in 25 mL of n-heptane with an initial DBT, 4,6-DMDBT, anthracene, naphthalene and xylene concentration of 20 ppm (0.1 mM), 22 ppm (0.1 mM), 180 ppm (1 mM), 130 ppm (1 mM), and 110 ppm (1 mM), respectively. Each suspension was stirred at a mixing speed of 250 rpm for 24 or 48 h. All adsorption studies were carried out at room temperature, and a temperature-controlled water bath was used to maintain the temperature constant. After 24 or 48 h, the solid was removed from the solution by centrifugation, and the UV-vis spectrum of the solution was obtained. To ensure the accuracy of the measurements and their reproducibility, adsorption studies were replicated three times, and the results were averaged and reported with a precision of ±1%.

Results and discussion

Preparation and characterization of 2a–2c for grafting on Si-MCM-41 mesoporous silica

In an exploratory experiment, the reaction of 1,5-dichloroanthraquinone (1a) with 3-aminopropyltriethoxysilane (3a) or 3-aminopropyltrimethoxysilane (3b) was performed in refluxing toluene. After completion of the reaction in 24 h (monitored by TLC) and subsequent workup, the products 1-chloro-5-((3-(triethoxysilyl)propyl)amino)anthracene-9,10-dione (2a) and chloro-5-((3-(trimethoxysilyl)propyl)amino)anthracene-9,10-dione (2d) were obtained as red solids in 65% and 67% yields, respectively. Initially, the product was assumed to be an imine derivative, as previously described concerning the reaction of 1-chloro-5-hydrazino-9,10-anthracendione with Boc-L-alanine.⁵⁴ However, a careful analysis of the NMR spectra of 2a–d did not support this assumption, specifically the ¹³C {¹H} NMR data. The symmetrical 1,5-dichloroanthraquinone (1a), 1,5-dicyanoanthraquinone (1b), and 1,5-dichloro-4,8-dinitroanthraquinone (1c) showed a single resonance at 180.9, 181.1 and 177.6 ppm, respectively, which can be attributed to anthraquinone carbonyl carbons. In the derivatives 2a–d, this symmetry is lost and two distinct anthraquinone carbonyl resonances are observed for amide products rather than only one resonance for imine products. The probable structures of 2a–d are shown in Scheme 1. The structure of 2d was confirmed by single-crystal X-ray analysis. The molecular structure of 2d is illustrated in Fig. 1 and its crystallographic data (Table S1†) along with a possible mechanism for the formation of 2a (Scheme S1†), which can be extended for 2b–d, are given in the ESI.†

Fig. 1 Molecular structure of 1-chloro-5-((3-(trimethoxysilyl)propyl)amino)anthracene-9,10-dione (2d) with hydrogen atoms and a labelling scheme.

The ¹H NMR spectra of 1-chloro-5-((3-(triethoxysilyl)propyl)amino)anthracene-9,10-dione (2a) exhibited a triplet for SiCH₂–CH₂–CH₂ at δ = 0.79, a triplet for the methyl group at δ = 1.25, a multiple for SiCH₂–CH₂–CH₂ at δ = 1.86–1.91, a quartet for SiCH₂–CH₂–CH₂–NH– at δ = 3.37, a quartet for OCH₂ at δ = 3.86, a doublet of doublet at δ = 7.05–7.09, a multiplet at δ = 7.53–7.73, a doublet of doublet at δ = 8.29–8.32 for aromatic protons, and broad peak at δ = 9.68 for the –NH of amine. The ¹H-decoupled ¹³C NMR spectrum of 2a showed 19 distinct resonances in agreement with the proposed structure. Finally, the molecular ion of 2a with the mass number of 463.22 m/z was observed in the mass spectrum.

Preparation and characterization of 4,6-dimethyldibenzothiophene:1,5-dichloro-4,8-dinitroanthraquinone charge-transfer complex

To study, understand and establish the type of interaction that takes place in the adsorption of DBT or 4,6-DMDBT by the electron-deficient-π-system-functionalized mesoporous silica, AQ-NO₂@Si-MCM-41, attempts were made to prepare a CTC with 4,6-DMDBT, 1,5-dichloroanthraquinone (1a), 1,5-dicyanoanthraquinone (1b), and 1,5-dichloro-4,8-dinitroanthraquinone (1c). All attempts to prepare CTCs only resulted in the formation of a CTC between 1,5-dichloro-4,8-dinitroanthraquinone (1c) and 4,6-DMDBT (Fig. 2). The fundamental reason for the formation of a CTC between 1,5-dichloro-4,8-dinitroanthraquinone and 4,6-DMDBT is the presence of four electron withdrawing groups on anthraquinone, which make it a highly electron deficient system and susceptible to interaction with 4,6-DMDBT, which is an electron rich system, through π–π interaction. Interestingly, an earlier report on the crystal structure of 1,5-dicyanoanthraquinone shows that adjacent molecules are stacked over each other in contrast to 1,5-dichloro-4,8-dinitroanthraquinone,⁴⁷ and this provides good potential for the formation of a CTC for the latter mentioned compound.

Fig. 2 Charge-transfer complex formation between 4,6-dimethyldibenzothiophene and 1,5-dichloro-4,8-dinitroanthraquinone.

The CTC was fully characterized via ¹H and ¹³C {¹H} NMR spectroscopy, mass spectrometry, infrared spectroscopy, powder X-ray diffraction, thermal and elemental analysis and single-crystal X-ray crystallography. ¹H and ¹³C NMR data show the presence of two components of the CTC in the expected ratio.

The TGA-DTA curves of the CTC are shown in Fig. 3. In the TGA curve of the CTC, there is distinct weight loss from 205 to 292 °C, which is accompanied by an intense endothermic peak in the DTA curve. This event is associated with the release of 4,6-dimethyldibenzothiophene (33.6% weight loss cf. calculated 36.9%). Interestingly, no event occurs around the reported melting point of 153–157 °C for 4,6-dimethydibenzothiophene, which indicates a strong interaction between the two components of the CTC. The second weight loss (56.5%) in the TGA curve in the 350–400 °C region, which is accompanied by an endothermic peak at the same temperature, is attributed to the decomposition of the 1,5-dichloro-4,8-dinitroanthraquinone (1c) component of the CTC. Finally, thermal analysis showed that the combustion of the residue of the decomposition product takes place at 550 °C. Interestingly, thermal analysis of pure 1c (Fig. S1†) showed a similar endothermic peak in its DTA curve but at a somewhat lower temperature. Observation of the endothermic peak in the DTA curve of 1c at a lower temperature along with the absence of a subsequent exothermic peak in the DTA curve indicates its involvement in the CTC formation.

Fig. 3 TGA and DTA curves of the charge transfer complex 4,6-dimethyldibenzothiophene:1,5-dichloro-4,8-dinitroanthraquinone.

To assure the phase purity of the CTC, simulated and experimental X-ray-powder diffraction patterns of the CTC were recorded (Fig. S2†), which were consistent and thus indicate the high phase purity of CTC.

Single-crystal X-ray analysis

Single-crystal X-ray structure analysis of CTC was carried out and its molecular structure was determined (Fig. 4). The asymmetric unit contains half a molecule of 1,5-dichloro-4,8-dinitroanthraquinone (1c) and half a molecule of 4,6-DMDBT. Indeed, the entire core of 1c in the CTC with a root-mean-square deviation (RMSD) of 0.005 Å indicates a planer structure. In contrast, the central ring in the free 1c with an RMSD deviation of 0.091 Å deviates from the plane form, since the dihedral angle between the outer rings is 7.95°.⁴⁸ In fact, the deviation from the planarity of the anthraquinone moiety in the CTC is minimized and 1,5-dichloro-4,8-dinitroanthraquinone for the formation of the CTC with 4,6-DMDBT changed its configuration to completely planer. The CTC crystal is composed of alternate layers of donors and acceptors so that the planes of both aromatic molecules are parallel with the interplanar distance of 3.53 Å. The single-crystal structure of CTC shows two rings of 4,6-DMDBT participating in a π–π interaction with three rings of the anthraquinone molecule. The crystallographic data for the CTC and 1c are listed in the ESI (Table S2†).

Fig. 4 Molecular structure of the charge transfer complex of 4,6-dimethyldibenzothiophene:1,5-dichloro-4,8-dinitroanthraquinone with hydrogen atoms and labelling scheme.

Characterization of anthracene-9,10-dione derivative functionalized Si-MCM-41

Scheme 2 shows the grafting of 1-chloro-5-((3-(triethoxysilyl)propyl)amino)anthracene-9,10-dione (2a, AQ-Cl), 9,10-dioxo-5-((3-(triethoxysilyl)propyl)amino)-9,10-dihydroanthracene-carbonitrile (2b, AQ-CN) and 1,5-dichloro-4-nitro-8-((3-(ethoxysilyl)propyl)amino)anthracene-9,10-dione (2c, AQ-NO₂) on Si-MCM-41 mesoporous silica. The reactions of 2a–c with hydroxyl groups on the surface of Si-MCM-41 were fast and facile. Since AQ-NO₂@Si-MCM-41 showed good performance for the adsorption of DBT and 4,6-DMDBT, it was fully characterized.

Scheme 2 Grafting of 1-chloro-5-((3-(triethoxysilyl)propyl)amino)anthracene-9,10-dione (2a), 9,10-dioxo-5-((3-(triethoxysilyl)propyl)amino)-9,10-dihydroanthracene-carbonitrile (2b) and 1,5-dichloro-4-nitro-8-((3-(ethoxysilyl)propyl)amino)anthracene-9,10-dione (2c) on Si-MCM-41 (X = Cl, Y = H; X = Cl, Y = NO₂ and X = CN, Y = H).

The FT-IR spectrum of AQ-NO₂@Si-MCM-41 showed characteristic bands of 1,5-dichloro-4-nitro-8-anthracene-9,10-dione at 3085, 1688, 1639, 1546, 1407, 1381, 1310, 1241, 1147, 825 and 729 cm⁻¹ along with Si-MCM-41 typical bands at 1085, 805 and 459 cm⁻¹. The FT-IR data demonstrate that an electron-deficient π-system was successfully grafted on the Si-MCM-41 mesoporous silica. This is in accordance with the ²⁹Si CP/MAS NMR spectrum, which shows a new ²⁹Si peak at −65.8 ppm for AQ-NO₂@Si-MCM-41 (Fig. S3†) along with the Si-MCM-41 original peak at −107.3 ppm. Finally, in the ¹³C CP/MAS NMR spectrum of AQ-NO₂@Si-MCM-41, peaks were observed at 11.4, 23.5, 35.6, 45.7, 65.7, 86.1 ppm for the aliphatic area and 101.2, 115.3, 135.2, 151.2, 165.8, 184.7, 201.2, 215.3, 233.2 ppm for the aromatic area carbons (Fig. S4†).

Small-angle powder X-ray diffraction patterns of AQ-NO₂@Si-MCM-41 and bare Si-MCM-41 were recorded (Fig. S5†). As the XRD pattern shows, the mesoporous (100) characteristic peak remains intact, and this indicates that the mesoporous structure of Si-MCM-41 is maintained after grafting. However, the intensity of this peak dropped significantly and reflects that the Si-MCM-41 pores are filled with AQ-NO₂ in addition to grafting on the surface.

Thermal analysis (TGA-DTA) was used to investigate the thermal stability and loading of the grafted material on Si-MCM-41 (Fig. S6†). The TGA curve shows that a 7% weight loss occurs upon heating to 230 °C, along with a small endothermic peak in the DTA curve in the same temperature range, which suggests that the weight loss is due primarily to the elimination of physically adsorbed water and possibly chemically bound water. At a higher temperature, two strong exothermic peaks centred at 320 and 500 °C were observed, which are accompanied by a 19% weight loss in the TGA curve. This weight loss is primarily associated with the combustion of grafted materials and clearly indicates that 1,5-dichloro-4-nitro-8-((3-(ethoxysilyl)propyl)amino)anthracene-9,10-dione (2c) is well grafted on Si-MCM-41. Finally, thermal analysis of two different AQ-NO₂@Si-MCM-41 samples gave a 0.22 and 0.10 mmol g⁻¹ loading of 1,5-dichloro-4-nitro-8-((3-(ethoxysilyl)propyl)amino)anthracene-9,10-dione (2c). As mentioned above, since AQ-NO₂@Si-MCM-41 showed good performance in the adsorption of DBT and 4,6-DMDBT, it was fully characterized.

Adsorption studies

The adsorption of DBT, 4,6-DMDBT, anthracene, naphthalene and xylene on Si-MCM-41, AQ-NO₂@Si-MCM-41, AQ-CN@Si-MCM-41 and AQ-Cl@Si-MCM-41 in heptane at room temperature was monitored by UV-vis spectroscopy (Fig. S7–S10†). As shown in Table 1, 0.10 g of AQ-NO₂@Si-MCM-41 with 0.1 mmol g⁻¹ AQ-NO₂ loading adsorbed 62% of 4,6-DMDBT that was present in the initial solution in 48 h, and it adsorbed 89% of DBT in the same period of time. It should be mentioned that the preparation of CTC of 1,5-dichloro-4,8-dinitroanthraquinone (1c) with 4,6-DMDBT in a solution was successful in contrast to the preparation of the CTC of 1c with DBT. This can be attributed to the electronic factor of 4,6-DMDBT, which enhances the formation of the CTC between 1c and 4,6-DMDBT in a solution. Interestingly, the ease of CTC formation between 1c and 4,6-DMDBT in a solution was well reflected in the adsorption of 4,6-DMDBT by AQ-NO₂@Si-MCM-41. An attempt to increase the adsorption of DBT and 4,6-DMDBT by AQ-NO₂@Si-MCM-41 with an increase in the loading of 1c on Si-MCM-41 was not successful. Notably, when the 1c loading increased from 0.10 mmol g⁻¹ to 0.22 mmol g⁻¹, the amount of DBT and 4,6-DMDBT adsorption dropped from 89% to 32% and 62% to 56%, respectively. It is most likely that in a high loading of AQ-NO₂ (1c) in MCM-41, steric hindrance, to some extent, interrupts the π–π interaction and this causes the adsorption of both aromatic sulphides to decrease. Moreover, a decline for DBT is somewhat higher. The reason for the lower decrease in the adsorption of 4,6-DMDBT with a higher loading of 1c is not well understood at this time. However, it seems that with a higher loading of 1c, the electronic factor, which reflects the electron donor ability of 4,6-DMDBT, overcomes the steric factor, and consequently, the amount of adsorption for 4,6-DMDBT experiences a smaller decrease. Finally, as Table 1 shows, an attempt to adsorb DBT or 4,6-DMDBT by AQ-Cl@Si-MCM-41 and AQ-CN@Si-MCM-41 in the same loading and conditions as described for AQ-NO₂@Si-MCM-41 was not successful. These results are in accordance with the failure in the CTC formation of 1,5-dichloroanthraquinone (1a) and 1,5-dicyanoanthraquinone (1b) with 4,6-DMDBT in solution, as described earlier. It seems most likely that the formation of a CTC between the species grafted on Si-MCM-41 and DBT or 4,6-DMDBT is a prerequisite for their adsorption on functionalized Si-MCM-41. As mentioned in the preparation of CTC, the presence of four electron-withdrawing groups on anthraquinone make it a highly electron deficient system and susceptible to interaction with 4,6-DMDBT through π–π interaction. Apparently, this interaction is reflected in the adsorption of 4,6-DMDBT on AQ-NO₂@Si-MCM-41. More interestingly, as shown in Table 1, when DBT and 4,6-DMDBT as adsorbates were replaced with aromatic hydrocarbons, such as anthracene, naphthalene and xylene, which are present in large amounts in gasoline and diesel oil, the sorbent did not show any tendency for their adsorption. This is in accordance with the molecular orbital calculations for the adsorption preference of aromatic organosulphur compounds over aromatic hydrocarbons by π-complexation.⁵⁵ The high preference of this sorbent for adsorbing DBT and 4,6-DMDBT demonstrates its good potential for application in real fuels. The high adsorption of aromatic sulphides in comparison with aromatic hydrocarbons can be possibly attributed to the distinct electron density difference between them and the tendency of the former towards π–π interactions. The high performance of AQ-NO₂@Si-MCM-41 in removing DBT (entry 9 in Table 1, 89%) and 4,6-DMDBT (entry 10 in Table 1, 62%) in comparison to earlier methods for the desulfurization is quite promising for industrial applications.

Table 1 Adsorption of DBT and 4,6-DMDBT by anthracene-9,10-dione derivatives functionalized Si-MCM-41

Entry	Sorbent	AQ loading (mmol g^−1)	Adsorbent^a	Desulfurization efficiency^b (%)
a Amount of sorbent: 0.1 g, time: 48 h.b The results are the average of adsorption efficiencies.
1	Si-MCM-41	0	DBT	0
2	Si-MCM-41	0	4,6-DMDBT	0
3	AQ-Cl@Si-MCM-41	0.22	DBT	<1
4	AQ-Cl@Si-MCM-41	0.22	4,6-DMDBT	<1
5	AQ-CN@Si-MCM-41	0.22	DBT	<1
6	AQ-CN@Si-MCM-41	0.22	4,6-DMDBT	<1
7	AQ-NO2@Si-MCM-41	0.22	DBT	32
8	AQ-NO2@Si-MCM-41	0.22	4,6-DMDBT	56
9	AQ-NO2@Si-MCM-41	0.10	DBT	89
10	AQ-NO2@Si-MCM-41	0.10	4,6-DMDBT	62
11	AQ-NO2@Si-MCM-41	0.10	Anthracene	<1
12	AQ-NO2@Si-MCM-41	0.10	Naphthalene	<1
13	AQ-NO2@Si-MCM-41	0.10	Xylene	<1

---

Fig. 5 shows the kinetics of DBT and 4,6-DMDBT adsorption on the AQ-NO₂@Si-MCM-41 adsorbent. The adsorption kinetics is higher for DBT than for 4,6-DMDBT in spite of the higher affinity of 4,6-DMDBT for the formation of CTC. Therefore, the slower adsorption of 4,6-DMDBT should be attributed to its higher steric barrier with two methyl substituents. Adsorption of the majority of DBT and 4,6-DMDBT takes place in less than 20 hours and for both substrates were nearly reached to a plateau in about 48 hours.

Fig. 5 Equilibrium isotherms for DBT and DMDBT adsorption on AQ-NO₂@Si-MCM-41 with 0.1 mmol g⁻¹ AQ-NO₂ loading under ambient condition.

Sorbent regeneration studies

A preliminary experiment showed that when a CTC of 4,6-dimethyldibenzothiophene:1,5-dichloro-4,8-dinitroanthraquinone is heated under reduced pressure at 200 °C, it generates colourless crystals of 4,6-DMDBT and a yellow powder. The X-ray powder diffraction analysis of the yellow material and comparison of its pattern with the simulated X-ray powder diffraction pattern of 1,5-dichloro-4,8-dinitroanthraquinone (1c) powder (Fig. S11†) confirmed the formation of CTC components. This encouraged us to investigate the regeneration of sorbents (entries 9 and 10 in Table 1). In this context, spent sorbents were heated under reduced pressure at 200 °C for 5 h and then cooled to room temperature and used for further desulfurization. Interestingly, the results showed that almost the original desulfurization capacity of sorbents can be achieved after several regeneration cycles and the kinetic of desulfurization are constant. Notably, no evidence was found for the release of the grafted anthraquinone with recycling, and this can provide the feasibility of reusing the sorbents many times. Furthermore, the regenerated sorbents were analysed via PXRD and the results indicate that the mesoporous structure of the sorbents remained intact and has the same long term stability as the fresh sorbent if regeneration is carried out below 250 °C. It is worth mentioning that in contrast to a majority of other sorbents,^37,41,45 the present materials can be regenerated by a simple mild heat treatment instead of using solvents, which makes these sorbents attractive and important from an economic perspective and environmental point of view.

Conclusion

In this study, the performance of a series of new adsorbents based the on the formation of a CTC among immobilized 9,10-anthraquinone derivatives on MCM-41 mesoporous silica and two aromatic sulphur compounds, DBT and 4,6-DMDBT, as well as anthracene, naphthalene and xylene was evaluated. The adsorptive desulfurization of DBT, 4,6-DMDBT, anthracene, naphthalene and xylene as models over the adsorbents was conducted under very mild conditions (low pressure, ambient temperature and no hydrogen consumption). Among the adsorbents, the adsorptive capacity of AQ-NO₂@Si-MCM-41 for DBT and 4,6-DMDBT is high. Since sorbent regeneration is a key element for commercial applications, this study presents that the AQ-NO₂@Si-MCM-41 sorbent can be successfully regenerated by a simple heat treatment under reduced pressure and reused several times without significant loss in its activity, which is a remarkable advantage of this new adsorbent. The results of this work provide insight into the ultra-deep desulfurization of liquid fuels. The experimental results also show that the desulfurization of fuel by AQ-NO₂@Si-MCM-41, based on π-complexation, can be carried out in the presence of light aromatic compounds with the selective adsorption of DBT and 4,6-DMDBT.

Acknowledgements

The financial support from the Iranian National Science Foundation (Grant No. 93026590) is gratefully acknowledged. We also thank the Institute of Materials Research and Engineering for providing solid-state NMR facility.

Notes and references

1. J. M. Beer, Prog. Energy Combust. Sci., 2000, 26, 301.
2. J. M. Beer, Technol. Forecast. Soc. Change, 2010, 77, 411.
3. P. S. Kulkarni and C. A. M. Afonso, Green Chem., 2010, 12, 1139.
4. P. De Filippis and M. Scarsells, Ind. Eng. Chem. Res., 2008, 47, 973.
5. C. Yansheng, L. Changping, J. Qingzhu, L. Qingshan, Y. Peifang, L. Xiumei and U. Welz-Biermann, Green Chem., 2011, 13, 1224.
6. J. Kou, C. Lu, W. Sun, L. Zhang and Z. Xu, ACS Sustainable Chem. Eng., 2015, 3, 3053.
7. V. C. Srivastava, RSC Adv., 2012, 2, 759.
8. Directive 2003/17/EC, 3 March 2003, Official Journal of the European Union.
9. P. Tan, Y. Jiang, X.-Q. Liu, D.-Y. Zhang and L.-B. Sun, ACS Sustainable Chem. Eng., 2016, 4, 2223.
10. Y. Wang, R. T. Yang and J. M. Heinzel, Ind. Eng. Chem. Res., 2009, 48, 142.
11. B. C. Gates, J. R. Katzer and O. C. A. Schutt, Chemistry of Catalytic Processes, McGraw Hill, New York, 1979.
12. H. LÜ, Y. Zhang, Z. Jiang and C. Li, Green Chem., 2010, 12, 1954.
13. I. A. Babich and J. A. Moulijn, Fuel, 2003, 82, 607.
14. N. Tang, Z. Jiang and C. Li, Green Chem., 2015, 17, 817.
15. A. Bösmann, L. Datsevich, A. Jess, A. Lauter, C. Schmitz and P. Wasserscheid, Chem. Commun., 2001, 2494.
16. S. Zhang and Z. C. Zhang, Green Chem., 2002, 4, 376.
17. J. Eßber, P. Wasserscheid and A. Jess, Green Chem., 2004, 6, 316.
18. P. De Filippis and M. Scarsella, Energy Fuels, 2003, 17, 1452.
19. S. Murata, K. Murata, K. Kidena and M. Nomura, Energy Fuels, 2004, 18, 116.
20. S. Aggarwal, I. A. Karimi and G. R. Ivan, Mol. BioSyst., 2013, 9, 2530.
21. Y. Sano, K. Sugahara, K.-H. Choi, Y. Korai and I. Mochida, Fuel, 2005, 84, 903.
22. A. S. H. Salem, Ind. Eng. Chem. Res., 1994, 33, 336.
23. Y. Shi, X. Zhang and G. Liu, ACS Sustainable Chem. Eng., 2015, 3, 2237.
24. C. Laborde-Boutet, G. Joly, A. Nicolaos, M. Thomas and P. Magnoux, Ind. Eng. Chem. Res., 2006, 45, 6758.
25. X. Yang, C. Cao, K. J. Klabunde, K. L. Hohn and L. E. Erickson, Ind. Eng. Chem. Res., 2007, 46, 4819.
26. J. H. Kim, X. Ma, A. Zhou and C. Song, Catal. Today, 2006, 111, 74.
27. M. Flytzani-Stephanopoulos, M. Sakbodin and Z. Wang, Science, 2006, 312, 1508.
28. H. Y. Huang, R. T. Yang, D. Chinn and C. L. Munson, Ind. Eng. Chem. Res., 2003, 42, 2427.
29. A. Takahashi, F. H. Yang and R. T. Yang, Ind. Eng. Chem. Res., 2002, 41, 2487.
30. R. T. Yang, A. Takahashi and F. H. Yang, Ind. Eng. Chem. Res., 2001, 40, 6236.
31. S. G. McKinley and R. J. Angelici, Chem. Commun., 2003, 2620.
32. A. J. Hernandez-Maldonado and R. T. Yang, AIChE J., 2004, 50, 791.
33. A. Takahashi, F. H. Yang and R. T. Yang, Ind. Eng. Chem. Res., 2000, 39, 3856.
34. A. J. Hernandez-Maldonado, S. D. Stamatis, R. T. Yang, A. Z. He and W. Cannella, Ind. Eng. Chem. Res., 2004, 43, 769.
35. A. J. Hernandez-Maldonado, R. T. Yang and W. Cannella, Ind. Eng. Chem. Res., 2004, 43, 6142.
36. A. Milenkovic, D. Loffreda, E. Schulz, H. Chermette, M. Lemaire and P. Sautet, Phys. Chem. Chem. Phys., 2004, 6, 1169.
37. X. Wei, S. M. Husson, M. Mello and D. Chinn, Ind. Eng. Chem. Res., 2008, 47, 4448.
38. R. Gil, J.-C. Fiaud, J.-C. Poulin and E. Schulz, Chem. Commun., 2003, 2234.
39. V. Meille, E. Schulz, M. Vrinat and M. Lemaire, Chem. Commun., 1998, 305.
40. A. Milenkovic, E. Schulz, V. Meille, D. Loffreda, M. Forissier, M. Vrinat, P. Sautet and M. Lemaire, Energy Fuels, 1999, 13, 881.
41. M. Sevignon, M. Macaud, A. Favre-Reguillon, J. Schulz, M. Rocault, R. Faure, M. Vrinat and M. Lemaire, Green Chem., 2005, 7, 413.
42. P. Seth, A. Bauza, A. Frontera, C. Massera, P. Gamez and A. Ghosh, CrystEngComm, 2013, 15, 3031.
43. J. W. G. Bloom and S. E. Wheeler, Angew. Chem., Int. Ed., 2011, 50, 7847.
44. S. Grimme, Angew. Chem., Int. Ed., 2008, 47, 3430.
45. A. Favre-Reguillon, M. Sevignon, M. Rocault, E. Schulz and M. Lemaire, Ind. Eng. Chem. Res., 2008, 47, 9617.
46. M. Armaghan, M. M. Amini, S. W. Ng and E. R. T. Tiekink, Z. Kristallogr., 2013, 228, 598.
47. D. D. Perrin, W. L. F. Armarego and D. R. Perrin, Purification of Laboratory Chemicals, Pergamon Press, New York, 2nd edn, 1980, p. 20.
48. M. Armaghan, M. M. Amini and S. W. Ng, Acta Crystallogr., Sect. E: Struct. Rep. Online, 2010, 66, o767.
49. M. Armaghan, M. M. Amini and S. W. Ng, Acta Crystallogr., Sect. E: Struct. Rep. Online, 2010, 66, o2130.
50. P. T. Tanev and T. J. Pinnavaia, Chem. Mater., 1996, 8, 2068.
51. SMART, APEX2, version 2.1, Bruker AXS Inc., Madison, Wisconsin, USA, 2005.
52. G. M. Sheldrick, SHELXS-97, University of Gottingen, Germany, 1997.
53. APEX2 and SAINT, Bruker AXS Inc., Madison, Wisconsin, USA, 2008.
54. A. Dikov, M. Dimitrova, I. Ivanov, R. Krieg and K. J. Halbhuber, Cell. Mol. Biol., 2000, 46, 1219.
55. A. Takahashi, F. H. Yang and R. T. Yang, Ind. Eng. Chem. Res., 2002, 41, 2487.

---

Footnotes

† Electronic supplementary information (ESI) available: Crystal data for 2d (Table S1), a proposed mechanism for the formation of 1-chloro-5-((3-(triethoxysilyl)propyl)amino)anthracene-9,10-dione (2a) (Scheme S1), TGA and DTA curves of 1,5-dichloro-4,8-dinitroanthraquinone (1c) (Fig. S1), experimental and simulated powder X-ray diffraction pattern (PXRD) for the charge-transfer complex (Fig. S2), crystal data for charge transfer complex (CTC) (Table S2), ²⁹Si CP/MAS NMR spectra of Si-MCM-41 and AQ-NO₂@Si-MCM-41 (Fig. S3), ¹³C CP/MAS NMR spectrum of AQ-NO₂@Si-MCM-41 (Fig. S4), X-ray diffraction patterns of Si-MCM-41 and AQ-NO₂@Si-MCM-41 (Fig. S5), TGA-DTA curves of AQ-NO₂@Si-MCM-41 (Fig. S6), UV-vis spectrum of DBT before adsorption and after adsorption on AQ-NO₂@Si-MCM-41, DBT before and after adsorption, on AQ-NO₂@Si-MCM-41 (Fig. S7), UV-vis spectrum of anthracene before and after contact with AQ-NO₂@Si-MCM-41 (Fig. S8), UV-vis spectrum of naphthalene before and after contact with AQ-NO₂@Si-MCM-41 (Fig. S9), UV-vis spectrum of xylene before and after contact with AQ-NO₂@Si-MCM-41 (Fig. S10) and simulated and experimental powder X-ray diffraction pattern for 1,5-dichloro-4,8-dinitroanthraquinone (1c) generated after heating CTC to 200 °C under reduced pressure (Fig. S11). CCDC 933691 and 1409263. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c6ra17453a

‡ Present address: College of Chemistry, Chemical Engineering and Materials Science Soochow University Suzhou, 215123, China.

---