The conversion of trans-ferulic acid to vanillin was achieved by heterogeneous catalysis. HKUST-1 was employed in this synthesis and the essential step in the catalytic process was the generation of unsaturated metal sites. Thus, when the catalyst was activated under vacuum, full vanillin conversion (yield of 95%) was obtained.
Vanillin (4-hydoxy-3-methoxybenzaldehyde) is the main component of the extract of vanilla bean. The natural vanilla scent is a mixture of approximately 200 different odorant compounds in addition to vanillin. The natural extraction of vanillin (from the orchid Vanilla planifolia, Vanilla tahitiensis and Vanilla pompon) represents only 1% of the worldwide production and since this process is expensive and very long, the rest of the production of vanillin is synthesized. Many biotechnological approaches can be used for the synthesis of vanillin from lignin, phenolic stilbenes, isoeugenol, eugenol, guaicol, etc., with the disadvantage of harming the environment since these processes use strong oxidizing agents and toxic solvents. Thus, eco-friendly alternatives on the production of vanillin are very desirable and thus, under current investigation. Porous coordination polymers (PCPs) are a new class of highly crystalline materials that recently have been used for catalysis. HKUST-1 (Cu3(BTC)2(H2O)3, BTC = 1,3,5-benzene-tricarboxylate) is a very well known PCP which has been extensively studied as a heterogeneous catalyst. Here, we report a synthetic strategy for the production of vanillin by the oxidation of trans-ferulic acid using HKUST-1 as a catalyst.
The use of porous coordination polymers (PCPs) as heterogeneous catalysts1-4 is a relatively new research field. Due to very interesting properties that PCPs show, e.g., porous regularity, high surface area and metal access, they can offer new alternatives for heterogeneous catalysts5-6. The generation of catalytically active PCPs has been the main focus of many research groups7-10. A porous coordination polymer is constituted by metal ions and organic linkers and thus, the catalytic activity of these materials is provided by any of these parts. Some PCPs contain unsaturated (active) metals that can catalyze a chemical reaction11. However, the generation of unsaturated metal sites (open metal sites) within coordination polymers is not a trivial task and it represents a synthetic challenge that can be summarized in: (i) the generation of vacant coordination by removal of labile ligands7-11; (ii) the generation of bimetallic PCPs by incorporating organometallic ligands (previously synthesized)8,12-13; (iii) the post-synthetic variation of the metal ions9,14-15 or to the organic ligands10, 16-17 within the pores of the PCPs. Since the methodology (i) is the simplest thus, it is the most frequently used. Typically, the generation of open metal sites has been used for enhancing the affinity of PCPs towards H218-19, as well as for designing active heterogeneous catalysts20-27. In order to achieve good catalyst properties, PCPs need to show, additionally to the accessibility of open metal sites, retention of the crystallinity after the catalytic experiment, relatively high thermal stability and chemical stability to the reaction conditions.
HKUST-1 (Cu3(BTC)2(H2O)3, BTC = 1,3,5-benzene-tricarboxylate)7 is a well-investigated porous coordination polymer constructed with Cu(II) cations, that are coordinated to the carboxylate ligands and water. Interestingly, these water molecules can be eliminated (by heating) and this provides a square planar coordination around the copper ions which exhibit hard Lewis acid properties11. Bordiga and co-workers28 showed that the elimination of these H2O molecules did not affect the crystallinity (retention of the regularity) and the oxidation state of the metal ions (Cu(II)) was not affected. The use of HKUST-1 as a catalyst has been extensively investigated29-33 and in particular (very relevant for the present work) the oxidation with hydrogen peroxide of aromatic molecules34.
Vanilla is one of the most widely used flavoring agents in the cosmetic, pharmaceutical and food industries. It is extracted from the cured beans of the orchid Vanilla planifolia, Vanilla tahitiensis and Vanilla pompon. The Mayan and Aztec civilizations (pre-Columbian people) first realized the enormous potential of vanilla as a flavoring agent since it improved the chocolate flavor35-37. Vanilla was first isolated in 185838 and it was not until 187439 that the chemical structure of vanillin was finally determined. The natural extraction of vanillin (from the orchid Vanilla planifolia, Vanilla tahitiensis and Vanilla pompon) represents only 1% of the worldwide production and since this process is expensive and very long40, the rest of vanillin is synthesized40. Many biotechnological approaches can be used for the synthesis of vanillin from lignin, phenolic stilbenes, isoeugenol, eugenol, guaicol, etc. However, these approaches have the disadvantage of harming the environment since these processes use strong oxidizing agents and toxic solvents41-43. Herein, we report a synthetic strategy for the production of vanillin by the oxidation of trans-ferulic acid using HKUST-1 as a catalyst.
CAUTION: The chemicals used in this catalytic procedure are relatively low in toxicity and non-carcinogenic. Please use all appropriate safety precautions when performing this experimental procedure such as safety glasses, gloves, lab coat, full length pants and closed-toe shoes. One part of the following procedures involves standard air-free handling techniques.
1. Activation of the Catalyst (HKUST-1)
2. Synthesis of Vanillin via Heterogeneous Catalysis
Three representative samples of HKUST-1 were analyzed by infrared spectroscopy: non-activated, activated at 100 °C for 1 hr in an oven (exposed to air), and activated under vacuum (10-2 bar) at 100 °C for 1 hr. Thus, Fourier transform infrared (FTIR) spectra were recorded using a spectrometer with a single reflection diamond ATR accessory (Figure 1). For all spectra, 64 scans in the 4,000 to 400 cm-1 range were recorded with a spectral resolution of 4 cm-1.
In order to confirm the composition of purified product from oxidation of trans-ferulic acid using HKUST-1 as a catalyst, a 1H-NMR spectra was performed on a NMR spectrometer under a static magnetic field of 11.74 T at 25 °C (Figure 2). For that, a 5 mg sample was introduced into NMR test tube and 0.5 ml of deuterated chloroform (CDCl3) was added.
Figure 1: FTIR spectra of the catalyst (HKUST-1) at 25 °C. Non-activated (green line), activated in a conventional oven (purple line) and activated under vacuum (orange line). Reproduced from reference44 with permission from the Centre National de la Recherche Scientifique (CNRS) and the Royal Society of Chemistry. Please click here to view a larger version of this figure.
Figure 2: 1H-NMR spectra of vanillin after column chromatography purification on silica gel when the catalyst (HKUST-1) was activated under vacuum (10-5 bar) and 100 °C for 1 hr.
The fundamental step for the catalytic conversion of trans-ferulic acid to vanillin was the activation of the catalyst (HKUST-1). If the catalyst is not activated in situ (under vacuum and at 100 °C), only partial conversion of trans-ferulic acid to vanillin was observed44. In other words, the accessibility to open metal sites is crucial for the catalytic cycle44, and this can be achieved by the elimination of coordinated water to the Cu(II) metal sites within the porous coordination polymer.
Therefore, in order to investigate this phenomenon three infrared experiments were carried out. The first experiment, the FTIR spectrum of a non-activated sample of HKUST-1 (Figure 1) showed the characteristic absorption bands at 3,400 cm-1 and 3,680 cm-1 corresponding to uncoordinated water (broad absorption band) and coordinated water (sharp absorption band), respectively. Second, the HKUST-1 sample activated in a conventional oven (exposed to air) showed a FTIR spectrum with minimal changes in comparison to the previous experiment: a decrease in the intensity of the absorption band at 3,400 cm-1 but no change on the sharp absorption band at 3,680 cm-1 (Figure 1). This change suggested an incomplete loss of uncoordinated water and the maintenance of coordinated water to the Cu(II) metal ion. Finally, the FTIR spectrum of the catalyst activated in situ exhibited two major changes (respect to the non-activated catalyst), a considerable reduction in the intensity of the absorption band at 3,400 cm-1 and a complete lost on the intensity of the absorption band at 3,680 cm-1. Thus, when the catalyst is activated under vacuum (10-2 bar) and at 100 °C, it is possible to get complete access to the Cu(II) open metal sites; the elimination of any water molecules (non and coordinated) inside the pores of HKUST-1.
Once the catalytic reaction is finished, the catalyst can be easily recovered by filtration and re-used (Figure 2)44. Of course, it has to be re-activated (under vacuum, 10-2 bar, and at 100 °C). In order to facilitate the re-activation process, once the catalyst is recovered, washing it up with approximately 200 ml of ethyl acetate is needed. Then, the catalyst can be left at room temperature (on the Buchner funnel) for 30 min, recovered and finally transferred into the activation system (vide supra).
The scaling up of the reaction is the current limitation of this experimental technique. The maximum amount of catalyst used was 0.05 g which resulted in a conversion yield of 95% (obtained after isolation of the vanillin by flash column chromatography)44. This catalytic reaction was carried out with higher quantities of catalyst (consequently, more ferulic acid and the rest of the chemicals) and the conversion yield dropped down significantly.
Only the combination of activated heterogeneous catalyst with hydrogen peroxide is needed to catalyze the oxidation of trans-ferulic acid. In order to discriminate the heterogeneous catalytic conversion, blank tests (one without HKUST-1 and another without H2O2) were performed44. After these reactions were completed, in the same conditions, the analysis of product composition did not show the presence of vanillin44.
The catalytic methodology presented in here showed how the heterogeneous catalyst (HKUST-1) can be used for the effective conversion of trans-ferulic acid to vanillin when it is activated in situ. Additionally the retention of the crystallinity of the catalyst was corroborated by PXRD44 after this was re-used. Previous synthetic vanillin methodologies use strong oxidizing agents and toxic solvents41-43. The current methodology completely avoids these hazards and it also offers the ability to re-use the heterogeneous catalyst (HKUST-1).
Finally, we are investigating the applicability of the catalytic technique for the oxidation of different α,β-unsaturated carboxylic acids.
The authors have nothing to disclose.
The authors thank Dr. A. Tejeda-Cruz (X-ray; IIM-UNAM). R.Y. thanks CINVESTAV, Mexico for technical support. M.S.S acknowledges the financial support by Spanish Government, MINECO (MAT2012-31127). I.A.I thanks CONACyT (212318) and PAPIIT UNAM (IN100415), Mexico for financial support. E.G-Z. thanks CONACyT (156801 and 236879), Mexico for financial support. Thanks to U. Winnberg (ITAM and ITESM) for scientific discussions.
HKUST-1 | Sigma-Aldrich | MFCD10567003 | |
Ferulic Acid (trans-4-Hydroxy-3-methoxycinnamic acid) | Sigma-Aldrich | 537-98-4 | |
Ethanol | Sigma-Aldrich | 64-17-5 | |
Hydrogen peroxide solution | Sigma-Aldrich | 7722-84-1 | |
Acetonitrile | Sigma-Aldrich | 75-05-8 | |
Ethyl acetate | Sigma-Aldrich | 141-78-6 | |
Ammonium chloride | Sigma-Aldrich | 12125-02-9 | |
Sodium sulfate anhydrous | Sigma-Aldrich | 7757-82-6 | |
Ethyl acetate | Sigma-Aldrich | 141-78-6 | |
n-Hexane | Sigma-Aldrich | 110-54-3 | |
Silica Gel | Sigma-Aldrich | 112926-00-8 | Size 70/230 |
250 mL two-neck round-bottom flask | Sigma-Aldrich | Z516872-1EA | 250 mL capacity |
Magnetic stirring bar | Bel-Art products | 371100002 | Teflon, octagon |
Condenser | Cole-Parmer | JZ-34706-00 | 200 mm Jacket length |
Vacuum pump (Approx. 10X-2 bar) | Cole-Parmer | JZ-78162-00 | Vacuum/Pressure Diaphragm Pump |
Stopcock | Cole-Parmer | EW-30600-00 | with a male luer slip |
Hose | Cole-Parmer | JZ-06602-04 | 16.0 mm ID and 23.2 mm ED |
Rubber septums | Cole-Parmer | JZ-08918-34 | Silicone with PTFE coating |
Hot plate | Cole-Parmer | JZ-04660-15 | 10.2 cm x 10.2 cm, 5 to 540 °C |
Sand bath | Cole-Parmer | GH-01184-00 | Fluidized Sand Bath SBS-4, 50 to 600 °C |
N2 gas | INFRA | Cod. 103 | Cylinder 9m ³ |
Ballons (filled with N2 gas) | Sigma-Aldrich | Z154989-100EA | Thick-wall, natural latex rubber |
Syringes with removable needles | Sigma-Aldrich | Z116912-100EA | 10 mL capacity |
Filter paper | Cole-Parmer | JZ-81050-24 | Grade No. 235 qualitative filter paper (90 mm diameter disc) |
Buchner funnel | Cole-Parmer | JZ-17815-04 | 320 mL capacity which accept standard paper filter sizes |
Buchner flask | Cole-Parmer | JZ-34557-02 | 250 mL capacity |
Rotary Evaporator | Cole-Parmer | JZ-28710-02 | |
Beakers | Cole-Parmer | JZ-34502-(02,04,05) | Pyrex Brand 1000 Griffin; 20, 50 and 100 mL |
Separation funnel | Cole-Parmer | JZ-34505-44 | Capacity for 125 mL with steam lenght of 60 mm |
Glass column for chromatography | Cole-Parmer | JZ-34695-42 | Column with fritted disk, 10.5 mm ID x 250 mm L |
PXRD diffractometer | Bruker | AXS D8 Advance XRD | |
FTIR spectrophotometer | Thermo scientific | FT-IR (JZ-83008-02); ATR (JZ-83008-26) | Nicolet iS5 FT-IR Spectrometer, with KBr Windows and iD5 Diamond ATR |