Magnetic silica beads functionalized with cobalt phthalocyanine for the oxidation of mercaptans in an alkali free aqueous medium

Abstract

Novel magnetic silica beads functionalized cobalt phthalocyanine catalyst was prepared by immobilization of the sulfonated Co(II) phthalocyanine (CoPcS) on the amino functionalized silica coated magnetic nanoparticles (Fe₃O₄@SiO₂, SMNP) via a sulfonamide linkage. The developed catalyst was characterized by various techniques such as XRD, N₂ adsorption–desorption, UV-Vis, FTIR, SEM and TGA analyses. The synthesized catalyst was found to be effective in oxidation of mercaptans in an aqueous medium by using molecular oxygen as oxidant under alkali free conditions. The efficient recovery by using an external magnet along with reusability of the catalyst in an alkali free reaction condition makes this methodology a facile, clean, effective and economically feasible approach for the oxidation of mercaptans to disulfides.

---

Mercaptans (RSH) are present in petroleum products like LPG, naphtha, gasoline, kerosene, and ATF. RSH are undesirable due to their foul odor, and corrosive nature.¹ They may also affect the activity of downstream catalysts. Therefore, it is necessary to convert them to a less deleterious form before end use. The process for mercaptan conversion is known as sweetening.^2,3 It involves the oxidation of RSH by air to disulfides in an alkaline medium in the presence of water soluble sulfonated phthalocyanines (CoPcS) as catalyst.^4–6 However, the use of caustic is disadvantageous because of the hazardous nature of the spent caustic and its disposal being expensive due to stringent environmental regulations.^7,8 Furthermore, homogeneous nature of the catalyst makes this process less attractive due to difficulty in recovery and reuse of the catalyst.⁹ Thus, a number of studies have been directed towards the improvement of cobalt phthalocyanine catalytic systems as well as immobilization of cobalt phthalocyanines on solid supports to make the process caustic free.^7,10 In particular, solid basic materials, such as magnesium oxides,^11–13 mixed Mg–Al oxides¹⁴ and surface-modified carbons^15,16 have been used for supporting these catalysts. Though solid basic catalysts are less stable, yet they have the potential to provide alkali free environment for oxidation of mercaptans to disulfides using heterogenized cobalt phthalocyanines as catalyst.

Recently, nanoparticles have emerged to be an efficient alternative for the immobilization of metal complexes to make them easily recoverable and recyclable. Nevertheless, difficult recovery of the nanosized particles from the reaction mixture limits their wide spread applications. Therefore, research focus has presently shifted instead to the use of magnetic nanoparticles (such as Fe₃O₄) coated with metal complexes which combines the advantages of simplistic recovery of catalyst using external magnet with the catalytic activity of the metal complexes.^17–21

In our ongoing research efforts to develop novel catalytic methodologies for organic transformations, we report herein an efficient and alkali free oxidation of mercaptans to disulfides in an aqueous medium. The reaction was catalyzed using a novel magnetically recoverable nanosized silica coated magnetic nanoparticles (ASMNP) immobilized tetrasulfonated Co(II) phthalocyanine (CoPcS) using molecular oxygen as an oxidant (Scheme 1). The catalyst was recovered and reused several times without significant loss in the catalytic activity.

Scheme 1 Oxidation of mercaptans to disulfide using magnetically separable CoPcS catalyst.

Synthesis and characterization of the catalyst

Magnetic nanoparticles (MNP) were prepared by co-precipitation method using FeCl₂ and FeCl₃ in aqueous hydrochloric acid at room temperature.²² The as-obtained particles were subsequently coated with a dense silica layer, using tetraethoxysilane (TEOS) as the silica source and ammonia solution. The silica bound magnetic nanoparticles (SMNP) were subsequently used for the immobilization of cobalt phthalocyanine by using (3-aminopropyl) triethoxysilane (APS) as the chelating ligand as shown in Scheme 2. In the synthesized catalyst, the coating of magnetic nanoparticles with a SiO₂ layer containing CoPcS attached to surface is critical, as it reduces the undesirable oxidation of nanomagnetite surface by oxygen and therefore improves its stability.

Scheme 2 Step wise synthesis of silica coated magnetic nanoparticles immobilized cobalt phthalocyanine (CoPcS@ASMNP).

The as-synthesized heterogeneous catalyst i.e. CoPcS@ASMNP was characterized by several techniques such as FT-IR, XRD, TGA, SEM, UV-Vis etc. The FT-IR spectra of MNP, ASMNP and the CoPcS@ASMNP are given in Fig. 1.

Fig. 1 FTIR spectra of: (a) MNP; (b) ASMNP; (c) CoPcS@ASMNP.

For the Fe₃O₄ nanoparticles, the characteristic absorption peaks at 589 and 492 cm⁻¹ are attributed to the Fe–O structure (Fig. 1A). Surface functionalization of Fe₃O₄ nanoparticles by APTES (3-aminopropyl triethoxysilane) is confirmed by the presence of Fe–O–Si bands at around 544.3 cm⁻¹. The band at 1072.2 is due to the stretching vibration of Si–O band and the two bands at 3425.3 and 1632 cm⁻¹ are due to the N–H stretching vibration and NH₂ bending mode of free NH₂ groups, confirm the existence of APTES (Fig. 1B).^23,24 The FTIR of the magnetically recoverable cobalt phthalocyanine [CoPcS@ASMNP] is shown in Fig. 1C. The covalent immobilization of the CoPcS on ASMNP was confirmed by the presence of characteristic peak at 1652 cm⁻¹ corresponding to sulfonamide (S=O)–NH bonds (Scheme 2). The peaks at about 3100 cm⁻¹ are due to the stretching vibration of the C–H on the aromatic ring, the peaks at about 1652 cm⁻¹ are due to the C=C stretching vibration on the aromatic ring, and the peak at 1498 cm⁻¹ indicates the C=N stretching vibration. The existence of this series of peaks shows the presence of the phthalocyanine ring. The peaks at 1100 cm⁻¹ are characteristic absorption peaks of the S=O symmetric and asymmetric stretching vibrations of the sulfonic acid groups and the peaks at 619 cm⁻¹ are the C–S stretching vibrations.

The crystalline nature of MNP was confirmed by XRD (Fig. 2A). A series of the characteristic peak for Fe₃O₄ nanoparticles (2θ = 30.10, 35.53, 43.08 and 57.10) were observed. These peaks are consistent with the standard pattern of Fe₃O₄ with a cubic structure. The similar XRD pattern of CoPcS@ASMNP (Fig. 2B), indicates that the surface modification did not change the crystal structure of the MNPs during the immobilization of CoPcS to ASMNP support.

Fig. 2 XRD spectra of (a) MNP; (b) CoPcS@ASMNP.

Furthermore, the synthesis of Fe₃O₄ nanoparticles was confirmed by XPS analysis (Fig. 3). The peaks at about 726.4 and 711.2 eV are attributed to the Fe²p_1/2 and Fe²p_3/2 respectively. These values are in accordance with the literature and are characteristic of Fe₃O₄ nanoparticles.²⁵

Fig. 3 XPS of magnetic nanoparticles (Fe₃O₄).

The SEM image of the CoPcS@ASMNP confirmed the uniformity and spherical morphology of the nanoparticles (Fig. 4) in the synthesized material. The size of the nanoparticles was found to be increased in the range of 50–100 nm, due to the aggregation of the nanoparticles. Further, EDX analysis confirmed the presence of cobalt and magnetic nanoparticles in the synthesized catalyst.

Fig. 4 SEM of (a) MNP; (b) ASMNP; (c) CoPcS@ ASMNP; (d) EDX of CoPcS@ASMNP.

The UV-Vis spectra of CoPcS and heterogeneous CoPcS@ASMNP are shown in Fig. 5. The UV-Vis spectra of heterogeneous catalyst exhibited two characteristic absorption bands of CoPcS e.g. at 330 nm (B-band) and 700 nm(Q band) respectively.^15,16

Fig. 5 UV-Vis spectra of: (a) CoPcS and (b) CoPcS@ASMNP.

The thermal behaviour of the catalyst was determined by the thermo-gravimetric analysis (TGA) under nitrogen atmosphere (Fig. 6). The decomposition of the catalyst was started at 250 °C and was completely decomposed between 250–500 °C, indicating the higher thermal stability of the catalyst. According to the TGA, the amount of residual cobalt after burning off phthalocyanine moieties was evaluated to be 5.5 wt%. Therefore each gram of heterogeneous catalyst contains 0.93 mmol g⁻¹ of Co or CoPcS complex. This value was found to be in good agreement with that of ICP-AES analysis (5.6 wt% Co, 0.95 mmol g⁻¹) catalyst.

Fig. 6 TGA of CoPcS@ASMNP.

The synthesized catalyst exhibited superparamagnetism, which is confirmed by the magnetization curve shown in Fig. 7 using a vibrating sample magnetometer (VSM). It can be seen that when magnetic field was applied, the material showed a strong response, with a saturation magnetization value of 4.42 emu g⁻¹.

Fig. 7 Magnetization curve of CoPcS@ASMNP.

Catalysis

1-hexanethiol was chosen as a representative substrate for preliminary studies. The oxidation of 1-hexanethiol in neat water under air did not proceed in the absence of catalyst at room temperature as well as at 50 °C. Similarly, the reaction did not occur by using silica coated magnetic nanoparticles as catalyst (0.1 g). On the other hand, the CoPcS@ASMNP catalyst (0.1 g) effectively catalyzed the oxidation of hexane thiol in water under alkali free conditions. The selectivity for corresponding disulfides was found to be almost 100% without any evidence for formation of other products being detected by GC. Importantly, the similar reaction was found to be very slow in the presence of homogeneous CoPcS catalyst under identical experimental conditions and produced only 20% yield of the disulfide (Table 1, entry 1). This is probably due to the proton acceptor tendency of the amino functionalized magnetic nanoparticles which facilitates the deprotonation of the mercaptan to give thiolate anion. Thus, the developed heterogeneous catalyst provided higher catalytic activity coupled with the benefits of heterogeneous catalyst such as facile recovery of the catalyst using an external magnet under alkali-free conditions.

Table 1 Oxidation of mercaptans to disulfides in an aqueous alkali free medium^a

Entry	Substrate	Product	Time/h	Yield^b (%)	TOF/h^−1
a Reaction conditions: mercaptan (2 mmol), catalyst (5 mol%; 0.1 mmol), water (3 ml) at 70 °C by purging molecular oxygen.b Isolated yield.c Using CoPcS as catalyst.d GC yield.
1			2.0	96	9.6
				20^c
2			2.5	93	7.4
3			2.0	96	9.6
4			3.5	92	5.2
5			5.5	60	2.2
6			3.0	92	6.1
7			3.0	89	5.9
8			3.0	87	5.8
9			3.5	85	4.8
10			4.5	82	3.6
11			6.5	78	2.4
12			8.0	70	1.7
13			8.0	67	1.7
14			10	60	1.2
15			10	55^d	1.1
16			12	42^d	0.7
17			12	35^d	0.6

---

The reaction was found to be slow in commonly used organic solvents such as ethyl acetate, dichloroethane and n-hexane, however it efficiently occurred in water and afforded complete conversion at room temperature. The use of water as a “green” reaction medium is a remarkable advantage, especially due to the very low solubility of the organic compounds in water, providing easy separation of the products.

The scope of the protocol was subsequently extended to a range of aliphatic and aromatic mercaptans, as shown in Table 1. All substrates could be smoothly converted to corresponding disulfides with high/excellent yields under mild conditions. The conversion of the reactions was determined by GC-MS and identity of the products was established by comparing the spectral data (¹H NMR) with the reported compounds (see ESI†). As shown in Table 1, aromatic mercaptans were found to be more reactive and provided better product yields than those of aliphatic mercaptans. Among the various thiophenols studied (Table 1, entries 1–5), those substituted with electron donating groups (Table 1, entries 2 and 3) were found to be more reactive than those having electron withdrawing groups (Table 1, entries 4 and 5). In the case of aliphatic mercaptans the reactivity decreases with the increase in chain length and accordingly the mercaptans with longer chain length require more reaction time. The possible reason for this reactivity difference may be the ease in abstraction of protons in the case of aromatic thiols compared to aliphatic thiols. This leads to production of corresponding thiolate anions which is the key to formation of disulphides and therefore responsible for higher reactivity of aromatic thiols as compared to aliphatic ones. Similarly in case of aliphatic thiols the abstraction of protons probably becomes difficult with increasing alkyl chain length leading to its decreased reactivity.

Furthermore, recycling of the catalyst was tested using the oxidation of thiophenol as a representative example. After completion of the reaction, the catalyst was easily separated by using external magnet. The catalyst was washed several times with water, dried and reused for subsequent experiments. The recovered catalyst was tested for ten subsequent runs under the optimized reaction conditions. The conversion and selectivity of the corresponding disulfide was found to be almost similar during these experiments (Table 2). These results established the facile recovery and efficient recycling of the catalyst. The recovered catalyst after ten cycles was characterized by using FTIR, XRD and ICP-AES analysis. The analyses were found to be in good agreement to the fresh one, confirming the stability as well as structural integrity of the catalyst (ESI Fig. S1 and S2†). Furthermore, the recovered heterogeneous catalyst after ten cycles showed 5.5 wt% of cobalt in ICP-AES analysis, indicating that there was no detectable leaching observed during the reaction cycles.

Table 2 Results on recycling experiments^a

Run	1	2	3	4	5	6	7	8	9	10
a Condition: thiophenol (2 mmol), catalyst (5 mol%, 0.1 mmol), water (3 ml) at 70 °C under oxygen atmosphere.b isolated yield.
Time/h	2.0	2.0	2.0	2.0	2.0	2.0	2.0	2.0	2.0	2.0
Yield^b (%)	96	95	95	94	94	94	94	95	94	94

---

Magnetic separation provides a very convenient approach for removing and recycling magnetic catalysts as shown in Fig. 8. After completion of the reaction, the catalyst was separated by placing a magnet near the glass bottle. The dark brown particles of CoPcS@ASMNP can be attracted towards the magnet with time increasing and was almost totally attracted after 5 min, leaving a clear solution.

Fig. 8 Separation of the catalyst by external magnetic effect.

Although the exact mechanism of the reaction is not clear at this stage. Based on the literature reports,^13,26 we assume that the amino functionalized silica coated magnetic nanoparticles owing to their proton acceptor properties may favor faster deprotonation of the substrate. The protons are subsequently captured for the oxidation of Co²⁺ into Co³⁺ by molecular oxygen to produce water. The oxidation of the thiolate anion (RS⁻) by Co³⁺ regenerates the catalyst and gives RS˙ radicals, which finally dimerises to give corresponding disulfide. As suggested in the literature,²⁷ the first step i.e. the deprotonation of thiol to thiolate ion is the key step of the whole process. Accordingly, it is presumed that the amino functionalization of MNPs is vital and facilitates the deprotonation of thiol to give thiolate anion.

In conclusion, a novel magnetically recoverable silica coated magnetic nanoparticles immobilized CoPcS catalyst with high thermostability (>250 °C) was synthesized for the oxidation of mercaptans in water under mild alkali free conditions. The oxidation of mercaptans was found to be very slow in the presence of homogeneous CoPcS complex as catalyst. On the other hand, the reaction rate was significantly enhanced in the presence of a catalytic amount of developed nanocatalyst and afforded high to excellent yield of the corresponding disulfide. The blank test without any catalyst at 70 °C did not give any oxidation product. After completion of the reaction, the catalyst was easily recovered by using external magnet and used for subsequent experiments. The catalyst was efficiently reused for ten runs and no leaching was observed. The facile recovery, efficient recycling along with the use of water as a standard “green” solvent, and alkali free reaction conditions, makes the developed methodology a cost effective and environmentally benign process for the oxidation of mercaptans to disulfides.

Techniques used

Structural morphology of the catalyst was determined by using Scanning electron microscopy (SEM) (Jeol Model JSM-6340F). X-ray diffraction pattern was executed using a Bruker D8 Advance diffractometer at 40 kV and 40 mA with Cu K_α radiation (λ = 0.15418 nm). Samples was taken in a glass slide and dried before analysis. Vibrational spectra (FT-IR) of samples were recorded on Perkin-Elmer spectrum RX-1 IR spectrophotometer. Thermal stability of synthesized catalyst was calculated by thermogravimetric analysis (TGA) with the help of a thermal analyzer TA-SDT Q-600. Sample were analyzed between temperature range was 40 to 900 °C with heating rate was 10 °C min⁻¹ under nitrogen flow. UV-visible absorption spectra for the synthesized catalyst were obtained using a Perkin Elmer lambda-19 UV-VIS-NIR spectrophotometer with a 10 mm quartz cell. Solid spectra were recorded by using BaSO₄ as a reference. Then the data were extracted by using software plotdigitalizer. XPS measurements were obtained on a KRATOS-AXIS 165 instrument equipped with dual aluminum–magnesium anodes using Mg K_α radiation (hυ = 1253.6 eV) operated at 5 kV and 15 mA with pass energy of 80 eV with an increment of 0.1 eV. To confirm the superparamagnetic nature of the catalyst, a vibrating sample magnetometer (VSM; Lakeshore, 7400) was used to obtain the magnetization curve of the nanoparticles. Cobalt content of catalyst in wt% was determined by inductively coupled plasma atomic emission spectrometer (ICP-AES, DRE, PS-3000UV, Leeman Labs Inc, USA). For preparation of sample 0.01 g of catalyst was digested with conc. HNO₃ and heated at 70 °C for 30 min for oxidizing organic component. Finally volume was made up to 10 ml by adding deionised water.

Synthesis of silica-coated magnetite nanoparticles (SMNP)

A solution containing FeCl₂ (2 g) and FeCl₃ (5.2 g) is added to deoxygenated (by purging nitrogen gas) 12.1 N HCl solution (25 ml). This solution was added very slowly under mechanical stirring to 250 ml of 1.5 M NaOH solution. An external magnet was used to separate the Fe₃O₄ nanoparticles from the solution, which subsequently washed with 0.01 N HCl (3 times).

The as synthesized magnetic nanoparticles (Fe₃O_4; 2 g) were suspended in 40 ml ethanol and 4 ml deionized water followed by sonication for 15 min. To the mixture, 1.5 ml of tetraethyl orthosilicate (TEOS) was slowly added and further sonicated for 10 min. Aqueous ammonia (25% wt% solution) was added to the solution under mechanical stirring. Thus obtained silica coated magnetic nanoparticles (Fe₃O₄@SiO₂, SMNP) were finally separated by an external magnet and washed thoroughly with ethanol: water and then dried under vacuum.

Functionalization of silica-coated magnetite nanoparticles (ASMNP)

Silica coated magnetic nanoparticles (SMNP, 1 g) were taken in toluene and added slowly to 3-aminopropyltriethoxysilane (APTES, 5 ml) under stirring. The resulting suspension was refluxed for 24 h. So obtained ASMNP material was separated by an external magnet and washed with methanol, dried in oven.

Immobilization of CoPcS (CoPcS@ASMNP)

Dried ASMNP (1 g) was mixed with water and the resulting suspension was subjected to ultrasonication for 10 min. After that tetrasulfonated cobalt phthalocyanine CoPcS (0.2 g) was added to the resulting suspension under mechanical stirring. The mixture was stirred for 2 h, the immobilized catalyst was seperated by external magnet and subsequently washed with water untill the filtrate became colorless. The loading of the catalyst was determined by ICP-AES analysis and was found to be 5.6 wt%, 59.70 μmol g⁻¹ of catalyst.

General procedure for the oxidation of thiols

To a stirred solution of mercaptan (2 mmol) in 5 ml of deionized water, CoPcS@ASMNP catalyst (5 mol%, 0.1 mmol) was added and the resulting mixture was stirred and heated at 70 °C under oxygen atmosphere for time as mentioned in Table 1. Progress of the reaction was monitored by TLC (SiO₂). At the end of the reaction the catalyst was separated by using an external magnet and the aqueous layer was extracted with diethyl ether. The combined organic layer was dried over anhydrous sodium sulfate and then concentrated under reduced pressure to give corresponding disulfide. Conversion of mercaptan to disulfide was determined by GCMS. The recovered catalyst was subsequently used for the recycling experiments.

Acknowledgements

We kindly acknowledge Director CSIR-IIP for his kind permission to publish these results. We are thankful to the Analytical Science Division of CSIR-IIP for providing help in the analyses of the samples.

Notes and references

1. B. Basu, S. Satapathy and A. K. Bhatnagar, Catal. Rev.: Sci. Eng., 1993, 35, 571.
2. G. D. Hobson, Modern Petroleum Technology, Part I, Wiley and Sons, New York, 1986.
3. J. R. Salazar, in Handbook of Petroleum Refining Process, ed. R. A. Meyers, McGraw-Hill, New York, 1986.
4. UOP Inc., Netherlands Appl. NL 7608824, February 13, 1978.
5. R. R. Frame, US Pat. 4374042, February 15, 1983.
6. D. H. J. Carlson and P. Urban, US Pat. 4018705, April 19, 1977.
7. J. J. Alcaraz, B. J. Arena, R. D. Gillespic and J. S. Holmgren, Catal. Today, 1998, 43, 89.
8. D. L. Holbrook, B. J. Arena, T. Verachtert and J. C. Bricker, Norwegian Radiation Protection Authority Annual Meeting, 1993.
9. I. W. C. E. Arends and R. A. Sheldon, Appl. Catal., A, 2001, 212, 175.
10. T. V. Rao, K. N. Rao, S. L. Jain and B. Sain, Synth. Commun., 2002, 32, 1151.
11. H. Hattori, Chem. Rev., 1995, 95, 537.
12. J. J. Alcaraz, B. J. Arena, R. D. Gillespie and J. S. Holmgren, Catal. Today, 1998, 43, 89.
13. D.-E. Jiang, G. Pan, B. Zhao, G. Ran, Y. Xie and E. Min, Appl. Catal., A, 2000, 201, 169.
14. (a) H. Liu, G. Ran, E. Min and X. Yang, J. Mol. Catal., 1998, 12, 221; (b) H. Liu, X. Yang, G. Ran and E. Min, Acta Phys.–Chim. Sin., 1999, 15, 918.
15. W. X. Chem, W. Y. Lu, Y. Y. Yao and M. H. Xu, Environ. Sci. Technol., 2007, 41, 6240.
16. V. N. Nemykin, A. E. Polshyna, S. A. Borisenkova and V. V. Strelko, J. Mol. Catal. A: Chem., 2007, 264, 103.
17. L. Menini, M. J. da Silva, M. F. F. Lelis, J. D. Fabris, R. M. Lago and E. V. Gusevskaya, Appl. Catal., A, 2004, 269, 117.
18. B. R. Vaddula, A. Saha, J. Leazer and R. S. Varma, Green Chem., 2012, 14, 2133.
19. L. Menini, M. C. Pereira, L. A. Parreira, J. D. Fabris and E. V. Gusevskaya, Appl. Catal., A, 2008, 254, 355.
20. M. J. Jacinto, O. H. C. F. Santos, R. F. Jardim, R. Landers and L. M. Rossi, Appl. Catal., A, 2009, 360, 177.
21. (a) M. Zhu and G. Diao, J. Phys. Chem. C, 2011, 115, 18923; (b) M. Zhu and G. Diao, J. Phys. Chem. C, 2011, 24743; (c) M. Zhu and G. Diao, Nanoscale, 2011, 3, 2748; (d) M. Zhu, C. Wang, D. Meng and G. Diao, J. Mater. Chem. A, 2013, 1, 2118.
22. Y. Kang, S. Risbud, J. Rabolt and P. Stroeve, Chem. Mater., 1996, 8, 2209.
23. Z. Zhang, F. Zhang, Q. Zhou, W. Zhao, B. Ma and Y. Ding, J. Colloid Interface Sci., 2011, 360, 189.
24. M. Masteri-Farahani and N. Tayyebi, J. Mol. Catal. A: Chem., 2011, 348, 83.
25. L. Bartolome, M. Imran, K. G. Lee, A. Sangalang, J. K. Ahnd and D. H. Kim, Green Chem., 2014, 16, 279.
26. G. Das, B. Sain, S. Kumar, M. O. Garg and G. Murali Dhar, Catal. Today, 2009, 141, 152.
27. L. Menini, M. C. Pereira, A. C. Ferreira, J. D. Fabris and E. V. Gusevskaya, Appl. Catal., A, 2011, 392, 151.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra02355b

‡ Presently working as a graduate student at CSIR-Central Building Research Institute (CBRI), Roorkee-247667, India

---