On-line version ISSN 0104-6632
Braz. J. Chem. Eng. vol. 15 no. 2 São Paulo June 1998
Aldol Condensation of Citral with Acetone oN Basic Solid Catalysts
C. NODA1, G. P. ALT1, R. M. WERNECK1, C. A. HENRIQUES2
and J. L. F. MONTEIRO1
1NUCAT - Núcleo de Catálise, COPPE/UFRJ, CP 68502, CEP: 21945-970,
Rio de Janeiro, RJ, Brazil Fax: (021) 290-6626
2 Instituto de Química, UERJ, Rio de Janeiro, RJ, Brazil
(Received: November 5, 1997; Accepted: February 3, 1998)
Abstract - The catalytic performance of solids with basic properties, such as CaO, MgO and hydrotalcites, was evaluated in the aldol condensation of citral and acetone, the first step in the synthesis of ionones from citral. The best results were obtained with CaO and hydrotalcite with high conversions (98%) and selectivities (close to 70% for the main product) observed for both of the catalyst. Such pseudoionone yields were greater than those reported in the literature for the homogeneous reaction.
Keywords: Citral, pseudoionones, basic catalysts, hydrotalcite, calcium oxide, magnesium oxide.
Aldol condensation of citral with acetone on basic catalysts and the subsequent cyclization of the intermediates over acid catalysts forming a and b -ionones are examples of processes involving products obtained from essentials oils used extensively in the perfumery and soap industries. b -ionone is also important as a precursor in the synthesis of vitamin A.
Ionones are commercially obtained through a conventional homogeneous catalytic process involving many operations. The first step consists in the obtention of pseudoionones using aldol condensation of citral with acetone catalyzed by KOH solution. Citral conversion is close to 60% and pseudoionone selectivity comes close to 80%. The second step is the obtention of a and b -ionones through the cyclization of pseudoionones in acid media. In this process conversion is almost 60%. a and b -Ionone selectivities depend on the strength of the acid used as catalyst. a -Ionone is obtained in the presence of weak acids, such as phosphoric acid. When a strong acid is used, for example, sulfuric acid, b -ionones are obtained. Operations such as neutralizations, washings, phase separations and vacuum distillations are associated with both steps.
The substitution of this conventional process by heterogeneous catalysis can reduce the consumption of reactants and the number of steps involved. For example, the neutralization and washing stages can be eliminated and substituted by a filtration stage. In addition, the possibility of selecting the morst active and/or the morst selective catalyst can increase the total yield of desired products and can also reduce by-product formation, thereby simplifying the distillation step.
This work studied the first step of the ionone synthesis process, the obtention of pseudoionones from citral and acetone by aldol condensation using basic solid catalysts, aiming at achieving a feasible substitution of classic homogeneous processes by heterogeneous ones. Pontentially, the latter are more selective and have simpler separation stages, thus reducing the costs of the process.
Citral was obtained by vacuum fractional distillation of Cuban lemongrass essential oil with 91.3% of the citral isomer mixture (geranial and neral). Acetone (99.5% purity) containing less than 0.007% water was purchased from VETEC.
The active forms of calcium and magnesium oxide were prepared according to Tsuji et al. (1994). Calcium and magnesium oxide precursors were dissolved in deionized water by stirring at room temperature for 24 hours and dried at 383 K for 24 hours. Finally, the active oxides (CaO and MgO) were obtained from their respective hydroxides by thermal treatment under vacuum up to 773 K.
The active form of hydrotalcite (HTC450) was obtained from the original material synthesized by Andrade (1997) by thermal decomposition in a flow of air up to 723 K. The chemical composition of the catalysts used was determined by atomic absorption (Perkin-Elmer AAS 1100B spectrophotometer) and the textural characteristics were determined by physical adsorption of N2 at 77 K (Micromeritics ASAP 2000). The water and carbon dioxide contents in these samples were determined by TGA/DTA (Rigaku TAS 1000). Samples of calcium oxide (CaO), magnesium oxide (MgO) and hydrotalcite (in the original form and after thermal treatment) were studied by DRX (type F Siemens Difractometer), with CuKa radiation, 35 KV and 30 mA. The LSUCRI (Least Square Unit Cell Refinement) program was employed for structural refinement.
The reaction was carried out in liquid phase in a stainless steel batch reactor with a 50 ml capacity. The reactor was magnetically stirred and heated with an oil bath. Each catalyst was preactivated ex-situ in a glass calcination cell. After activation, it was rapidly transferred to the reactor and a mixture of citral and acetone (1:1 molar) was immediatly added. The catalyst concentration was always 2% w/w.
The tests with hydrotalcite and magnesium oxide as catalyst were done at 330 and 398 K under autogenous pressure for 4 hours. When calcium oxide was used as catalyst the reaction was studied for 1, 2 and 4 hours at 330, 363, 383 and 398 K. At the end of each run, the system was cooled to room temperature and the two phases separated by decantation. The liquid phase was stored in a dark flask and kept in a freezer. For the chromatographic analysis (VARIAN 3700 Chromatograph, 35 m Carbowax 20 M capillary column and FID) it was dissolved in n-hexane (purity of 90% v/v). Gas chromatography mass spectrometry (Shimatzu QP 2000 A GC/MS) was also used for the identification of the reaction products.
RESULTS AND DISCUSSION
The Mg/Al ratio of the original hydrotalcite sample was 2.7. This value and the results of the TGA/DTA analysis allowed the following composition to be established for this sample:
For a calcium oxide sample prepared as described above and stored in a closed flask with no special precautions, atomic absorption analyses gave a calcium content lower than that anticipated for pure CaO. DRX analyses showed that this sample was a mixture of calcium carbonate and calcium hydroxide. The absence of a calcium oxide phase was attributed to the fast rehydratation and carbonation of the sample upon being exposed to ambient air.
Progressive carbonation of the original calcium hydroxide was also observed, since the amount of calcium carbonate in this sample increased with time up to a limiting value of about 40 wt% CaCO3. The so-called calcium oxide obtained by the decomposition of this sample was indeed a mixture of about 53 wt% CaO and 47 wt% CaCO3. This was the sample used in the catalytic runs.
For the magnesium hydroxide no carbonation was observed when the sample was treated and stored under conditions similar to those described above for Ca(OH)2. So, upon in-situ calcination of this sample, pure magnesium oxide was generated and used as catalyst. DRX analyses indicated that this sample suffered a slow rehydratation process upon being exposed to ambient air.
Table 1 shows the results of the textural characterization of the samples. It was observed that the BET area of magnesium oxide is four times larger than that of calcium oxide. This result could be attributed to a pore obstruction caused by the growth of a superficial phase of calcium carbonate. The inspection of the magnesium oxide isotherm suggests that the high mesoporosity of this material is due to intercrystalline voids. No microporosity was observed for either material.
When the results of the textural analysis of the calcined hydrotalcite were compared to those of the magnesium oxide, a lower BET area and a lower mesoporosity were observed. The mesoporosity of this material can also be attributed to intercrystalline spaces.
When the aldol condensation of citral with acetone was studied at 330 K, the conversion of citral was not significant. The highest values for citral conversion were obtained at 398 K, particularly for the calcium oxide (CaO) and the calcined hydrotalcite (HTC450) catalysts. When magnesium oxide (MgO) was employed as the catalyst, citral conversion was about 20%. These results, associated with those from Noda et al. (1997) for the self-condensation of acetone, indicate that the aldol condensation of citral with acetone requires stronger basic sites and higher temperatures than the self-condensation of acetone. This derives from the fact that when the acetone carbanion attacks the citral molecule forming a new carbon-carbon bond with the carbonylic carbon of citral, as seen in Scheme 2, the resonance stability of the citral molecule is lost.
For the three catalysts the main products of the aldol condensation of citral with acetone at 398 K were the pseudoionone isomers. The formation of superior condensation products. as a result of the reaction between pseudoionone and a molecule of acetone, was also observed. Selectivities for the main products (pseudoionones and the superior condensation products) at 98% conversion of citral are compared in Figure 2.
(a) t-plot method; (b) BJH method
Scheme 1: Obtention of ionones from citral.
Scheme 2: Acetone carbanion formation and reaction with the citral molecule.
Figure 1: Aldol condensation of citral with acetone at 398 K, 4 hours and 3 atm on the different catalysts.
Figure 2: Product selectivities for the aldol condensation of citral with acetone at 398 K, 4 hours and 3 atm on calcium oxide and hydrotalcite. PSI= Pseudoionones; PCS= Superior condensation products.
For the HTC450 catalyst pseudoionone (PSI) selectivity was higher than 70% and that of the superior condensation products (PCS) was lower than 10%. The use of calcium oxide as catalyst increased PCS selectivity to 15%.
The effect of temperature and reaction time on the reaction was evaluated for the calcium oxide catalyst. Figure 3 shows the variation in the conversion of citral with temperature and reaction time, while the results obtained for selectivity and yield of PSI are presented in Figures 4 and 5, respectively. As expected, reaction time presented a more significant influence at lower temperatures since total conversion was attained after two hours at 398 K. For a reaction time of 1 hour, increasing temperature caused a slight increase in PSI selectivity but a strong increase in citral conversion. For larger reaction times, temperature hardly affected PSI selectivity, which reached values between 60 and 70%.
PSI yields as a function of temperature at different reaction times are presented in Figure 5. It can be observed that at the highest temperature (398 K) the PSI yields achieved in this work were greater than those reported in the literature (Noda et all., 1987) for the reaction in homogeneous medium (48%).
Figure 3: Influence of temperature and reaction time on citral conversion using calcium oxide as catalyst.
Figure 4: Influence of temperature and reaction time on the pseudoionone selectivity using calcium oxide as catalyst.
Figure 5: Influence of temperature and reaction time on pseudoionone yield using calcium oxide as catalyst.
The aldol condensation of citral with acetone on basic solid catalysts was studied. The best results were obtained using calcium oxide and calcined hydrotalcite as catalysts. Citral conversion at 398 K and 3 atm was about 98% and pseudoionone selectivity was greater than 68%. Pseudoionone yield was higher than that reported in the literature for the homogeneous process, confirming the success of the substitution of the homogeneous catalyst by a heterogeneous one.
The effect of temperature and reaction time was also studied with calcium oxide as catalyst, and it was observed that both citral conversion and pseudoionone yield were favored by an increase in these parameters.
The reaction products were confirmed by CG/MS analyses and they were in accordance with the reaction mechanism proposed in the literature.
H.Tsuji, F. Yagi, H. Hattori and H. Kita, "Self-condensation of n-butiraldehyde over solid base catalysts" J. of catal, 148, 759-770 (1994). [ Links ]
Andrade, R.C.S., Masters thesis, UNICAMP, Campinas, SP (1997). [ Links ]
Noda, C., Masters thesis, PEQ/COPPE/UFRJ, Rio de Janeiro, RJ (1997). [ Links ]
C. Noda,; J. Magraner and P.J. Diaz, Rev. Cubana de Química (1987). [ Links ]