Synthesis and application of fatty acid derived templates for the preparation of mesostructured silica material

Abstract

Bio-Organic-Salts (BOSs) synthesized from fatty acid methyl esters (FAMESs) have been developed for use as a bio-derived template for mesostructured material. Fatty acid derived imidazolium salts were prepared from very high oleic acid (VHOSO), linseed oil, and castor oil by a three-step synthesis involving transesterification, amidation and quaternisation. The as prepared BOSs were successfully applied as structuring agents for hexagonal mesoporous materials without necessity of hydrothermal treatment, as confirmed by TEM, SAXS and nitrogen physisorption.

---

Growing concerns about global warming, and the depletion of fossil fuels is driving academic and industrial research in investigating the use of alternative renewable feedstock for application as fuels and chemicals. In terms of the latter, the US department of energy has identified in 2004 a large range of compounds that can be synthetized from biomass. The “Top Value Added Chemicals from Biomass” show the possibility to produce a huge number of building blocks and biomolecules from biomass resources, notably from starch, cellulose, hemicellulose and lignin, to generate chemical intermediates with application in various fields like polymers, textiles, pharmaceuticals and so on.^1,2 Depending on the raw material, various biorefinery concepts can be developed. Among them, the oilseed biorefinery has become quite popular. The bio-oils extracted from oilseeds are a valuable feedstock, notably for the production of biodiesel, glycerine based compounds, lubricants, cosmetics, plastics.^3–5 Within this context, we aim at adding another utility of the oilseeds: the application of fatty acid derivatives as template for the preparation of mesostructured materials. Indeed, mesostructured and mesoporous materials have become an important class of material in chemistry with promising future for different applications.⁶ However, these materials use an ionic or non-ionic template produced either from petroleum resources or via a multistep process involving complex multistep synthesis. For example, the standard template used for the preparation of hierarchical material MCM-41 is cetyltrimethylammonium bromide (CTAB), a quaternary ammonium salt composed of an amphiphilic head and a long hydrophobic chain, which is obtained from hexadecanol via halogenation with bromium and subsequent amination with trimethylamine. Since the last decade, some research groups studied the possibility to use ionic liquids as structuring agent for the preparation of mesoporous and mesostructured material. Indeed, ionic liquid are organic salts where the polar head is composed of imidazolium or pyrrolidinium salt (quaternary salt) with an alkyl chain that can be more or less longer to tune the amphiphily of the ionic liquid.^7,8 Imidazolium salts can form mesostructured silica material or mesoporous silica material via two different approaches: the hydrothermal method and the nanocasting method respectively. The first method simply plays on the concentration of the template in water to generate micelles and also different lyotropic phases (lamellar, rod-like or cubic phase), depending on the experimental conditions.^9,10 Self-organization can be also obtained via the nanocasting method. In this case no water is used but only pure ionic liquid with the silica precursor and a small amount of acid are mixed together. Hereby, the imidazolium rings generate π–π-stacking interactions with the neighbouring imidazolium ring enabling thus the formation of the wormlike mesoporous silica.^11,12 Herewithin we give the proof of concept for the use of vegetable oil derived templating agents for the synthesis of highly ordered mesoporous materials. Three vegetable oils were employed: Very High Oleic Sunflower Oil (VHOSO) linseed oil and castor oil. The compositions of these bio-oils are summarized in the Table 1. Two of the bio-oils, namely sunflower oil and castor oil (ricin), are mainly composed of mono-unsaturated fatty acids: C18:1 and C18:1-OH, respectively. On the other hand, linseed oil exhibits large amounts of C18:3, C18:2 and C18:1. The bio-organic salts were prepared in a multi-step synthesis (Fig. 1). In the first step, the bio-oils were transesterified with methanol to the corresponding fatty acid methyl ester (FAME) using classical alkaline conditions. After purification, the FAMEs were modified by amidation with 1-(3-aminopropyl)imidazole. Aminolysis of the FAME was catalysed by 1,5,7-triazabicyclo[4.4.0]dec-5-ene (TBD) under solvent-free conditions. Conversion of FAME to fatty imidazole was almost complete (94%–95%) for all composition studied. After several washing (removal of excess 1-(3-aminopropyl)imidazole), quasi-pure products were obtained. ¹H NMR and infrared analysis of the different fatty imidazole molecules confirmed the presence of the imidazole ring, connected to the long alkyl chain via the amide bond (Fig. S1 and S2†). The last step for preparing the templating agent involved the quaternization of the imidazole ring with 1-chloropropane to generate the corresponding salt. 1-Chloropropane was chosen since it is easy to handle (liquid component at room temperature) and the short chain does not affect the polarity of the head. Using an excess of 1-chloropropane ensured the total conversion in neat condition. The excess of 1-chloropropane was easily removed by evaporation and can be recycled if necessary. ¹H NMR analysis of the different BOS confirmed the formation of the imidazolium salts, recognized from the chemical shift of the protons (Fig. S3†). Since the different synthetized BOS showed good purity, no further purification was employed, thus minimizing the synthesis steps in agreement with principles of green chemistry. The use of the as obtained bio-organic salts for the preparation of mesostructured silica materials was studied applying the hydrothermal approach, in basic media (NH₃), inspired from Zhu et al. using different BOS/TEOS ratios.¹⁰ The as-prepared samples were denominated meso-BOS-X where BOS represents the fatty imidazolium from different oil sources and X the BOS/TEOS ratio. After preparation of the silica material, TGA analysis (under air) was performed to determine the ideal calcination temperature which was thus fixed to 550 °C (Fig. S4†). Fig. 2 shows the N₂-adsorption–desorption isotherms for each silica-BOS-X synthetized. The three materials (meso-oleic-0.125, meso-linseed-0.15 and meso-ricin-0.2) exhibit isotherms comparable to MCM-41 material,¹³ namely a typical type IV isotherm according to the IUPAC nomenclature. The absence of an hysteresis for meso-oleic, meso-linseed and meso-ricin indicates a narrow pore distribution as found from the BJH model (Fig. S5 and S6†). Further information about the structure was obtained by SAXS. The intensity profiles for all silica materials using the different BOS with various BOS/TEOS in a range between 0.05 and 0.2 are shown in Fig. S7.† As one can see from Fig. 3, in the optimal conditions, mesoporous materials were obtained for BOS-oleic-0.125 and BOS-linseed-0.15 with hexagonal p6mm structure. However, different degree of structuration were observed: the silica based on BOS-oleic shows a higher degree of structuration than the silica of BOS-linseed. Indeed, linseed oil is a mixture of three main fatty acid (C18:3, C18:2 and C18:1) which can degrade the self-organization of the template in water. However π–π stacking interactions with the neighbouring imidazolium rings, plus the hydrogen bonding interaction with the neighbouring amide group can enhance the self-organization of the different fatty imidazolium produced from the linseed oil providing a more stable two-dimensional crystal network, as observed for 12-HSA (12-HydroxyStearic Acid) organogel derived amide.¹⁴ For meso-oleic and meso-linseed, the BOS/TEOS molar ratio had significant impact, since a shift from the hexagonal to a lamellar self-organization was stated when the BOS/TEOS molar ratio overpassed 0.15. TEM pictures (Fig. 4) confirm the organization of the materials with parallel channels in a 2D hexagonal structure. The specific surface area (S.A) for the both material was found very high reaching 1594 m² g⁻¹ in case of meso-oleic-0.125, which is in the range of standard MCM-41 (800 to 1400 m² g⁻¹, depending of the synthesis condition).^15,16 In the case of meso-ricin-0.2, the structure were closer to a lamellar phase with a mesoporous network (Fig. 2 and 3). One can thus conclude that in our case the presence of a hydroxyl group in the middle of the alkyl chain in the template pertubates the self-assembly in aqueous media, contrary to what was reported for 12-HSA.^17,18 Consequently, meso-ricin has a disordered mesostructure whereby any space group or unit cell can't be defined. However, the material exhibit high surface area (between 1367 m² g⁻¹) with narrow pore diameter distribution around 2.8 nm (Table 2). TEM (Fig. 4) confirmed the results from SAXS and nitrogen physisorption, namely a narrow distribution of mesopores without any organisation in the silica framework.

Table 1 Composition of the bio-oils used

Fatty Acids/%	VHOSO	Castor oil	Linseed
C16:0	3.6	0.9	5.6
C18:0	3.0	1	4.2
C18:1	85.6	2.6	23.0
C18:2	6.0	3.6	15.8
C18:1 OH		90.6
C18:3	0.1		50.8
C20:0	0.3	0.5	0.1
C20:1	0.3	0.4	0.2
C22:0	0.8	0.4	0.2
C24:0	0.3		0.1

---

Fig. 1 Reaction for the preparation of bio-organic salt from fatty acid methyl esters (FAMEs).

Fig. 2 N₂-adsorption–desorption isotherms for the different meso-BOS synthetized after calcination.

Fig. 3 SAXS Intensity profiles of the mesoporous materials prepared with BOS as template; meso-BOS-X where BOS represents the fatty imidazolium from different oil sources, and X represent the BOS/TEOS molar ratio; synthesis at room temperature without hydrothermal treatment.

Fig. 4 TEM pictures of the meso-BOS-X synthetized in the optimal (A)/(B)-meso-oleic-0.125, (C)/(D)-meso-linseed.0.15, (E)/(F)-meso-ricin-0.2, (meso-BOS-X, where X represent the BOS/TEOS molar ratio).

Table 2 Textural and structural properties of the mesostructured silica material prepared with bio-organic salt as template

Meso-BOS-X	S.A/m^2 g^−1	V p/cm^3 g^−1	W BJH/nm	d spacing^a/nm	Unit cell parameter^b/nm	Wall thickness^c/nm
a From Braggs law using (100) plane. b a 0 = (2d100)/√3. c a 0 − WBJH; WBJH from desorption branch; X represent the BOS/TEOS molar ratio.
Oleic-0.125	1594	1.4	3.2	4.24	4.9	1.7
Linseed-0.15	1079	0.93	3.2	4.29	4.96	1.76
Ricin-0.2	1367	0.97	2.8	—	—	—

---

Finally, in order to confirm the stability of the template in basic media and more specifically evaluating the stability of the amide bond, we prepared meso-oleic-0.125 including a hydrothermal treatment (HT) step with a temperature between 80 °C and 130 °C for one day. From the result (see ESI file†), it clearly appears that the amide bond is retained no matter the employed hydrothermal temperature. N₂-adsorption–desorption show a typical type IV isotherm with a slight hysteresis (Fig. S8A and B†).

Nevertheless, the pore distribution profile suggested a decrease of the organization with increasing HT temperature. However, the SAXS patterns (Fig. S8C†) confirm that the 2D hexagonal structure is preserved and the degree of crystallite is only slightly affected, which was further evidenced by TEM (Fig. S9†).

In conclusion, the three bio-oils (VHOSO, linseed and castor oil) have been modified by amide coupling to an imidazole ring and quaternisation reaction. The as-obtained BOS showed different kinds of self-organization in aqueous media depending on the hydrophobic chain. BOS-oleic and BOS-linseed formed rod-like micelles under optimal conditions, whereby a higher degree of structuration was observed for BOS-oleic. The mixture of different hydrophobic chains in linseed oil is supposed to be the main factor for the less resolved structure in that case. On the other hand, BOS-ricin formed a mesoporous silica material of worm-like structure with regular pores, which was ascribed to the presence of the hydroxyl groups in the middle of the hydrophobic chain leading thus to another organization in water media. In comparison to the standard CTAB surfactant, the BOS-oleic are more suitable and closer to the concept of green chemistry (renewable feedstock, less reaction steps, less solvents), giving silica material with high specific area and good degree of structuration.

Acknowledgements

This work was performed, in partnership with the SAS PIVERT, within the frame of the French Institute for the Energy Transition (http://www.institut-pivert.com) selected as an investment for the Future. The French Government under the reference ANR-001-01 supported this work, as part of the investments for the Future. Financial support from Région Nord Pas de Calais and European FEDER for SAXS laboratory equipment and TEM facility are also gratefully acknowledged.

Notes and references

1. T. P. Werpy, G., 2004-08-01.
2. H. J. E. B. J. J. W. J. F. Johnson. D, 2007-10-01.
3. F. Beaudoin, O. Sayanova, R. P. Haslam, I. Bancroft and J. A. Napier, OCL: Ol., Corps Gras, Lipides, 2014, 21, D606.
4. R. Armangaud-Batut, OCL: Ol., Corps Gras, Lipides, 2006, 13, 187–189.
5. L. C. Meher, C. P. Churamani, M. Arif, Z. Ahmed and S. N. Naik, Renewable Sustainable Energy Rev., 2013, 26, 397–407.
6. M. E. Davis, Nature, 2002, 417, 813–821.
7. Z. Ma, J. Yu and S. Dai, Adv. Mater., 2010, 22, 261–285.
8. S. Polarz and M. Antonietti, Chem. Commun., 2002, 2593–2604.
9. T. Wang, H. Kaper, M. Antonietti and B. Smarsly, Langmuir, 2007, 23, 1489–1495.
10. K. Zhu, F. Požgan, L. D'Souza and R. M. Richards, Microporous Mesoporous Mater., 2006, 91, 40–46.
11. Y. Zhou and M. Antonietti, Chem. Mater., 2004, 16, 544–550.
12. Y. Zhou, J. H. Schattka and M. Antonietti, Nano Lett., 2004, 4, 477–481.
13. M. Grün, K. K. Unger, A. Matsumoto and K. Tsutsumi, Microporous Mesoporous Mater., 1999, 27, 207–216.
14. J. Toro-Vazquez, J. Morales-Rueda, V. A. Mallia and R. Weiss, Food Biophys., 2010, 5, 193–202.
15. B. Gouze, J. Cambedouzou, S. Parrès-Maynadié and D. Rébiscoul, Microporous Mesoporous Mater., 2014, 183, 168–176.
16. H. I. Meléndez-Ortiz, L. A. García-Cerda, Y. Olivares-Maldonado, G. Castruita, J. A. Mercado-Silva and Y. A. Perera-Mercado, Ceram. Int., 2012, 38, 6353–6358.
17. V. A. Mallia, M. George, D. L. Blair and R. G. Weiss, Langmuir, 2009, 25, 8615–8625.
18. V. A. Mallia and R. G. Weiss, J. Phys. Org. Chem., 2014, 27, 310–315.

---

Footnote

† Electronic supplementary information (ESI) available: ¹H NMR Infrared, porosimetry data, SAXS, TEM. See DOI: 10.1039/c5ra19495d

---