Ind. Eng. Chem. Res. 2005, 44, 9561-9564
9561
Reaction Control of the Robinson Annelation by Pressure and Temperature Manipulation under Supercritical CO2 Conditions Hajime Kawanami,*,† Keiichiro Matsushima,‡ and Yutaka Ikushima† Supercritical Fluid Research Center, National Institute of Advanced Science and Technology, 4-2-1 Nigatake, Miyagino-ku, Sendai 983-8551, Japan, and Hokkaido Industrial Research Institute, N19-S11 Kita-ku, Sapporo 060-0819, Japan
Robinson annelation under supercritical CO2 in the presence of various metal oxide catalysts without any organic solvent was studied to synthesize fused polycyclic compounds such as the series of steroid compounds. Under constant pressure and temperature, the reaction would have a limitation in the yield of desired product 8a-methyl-3,4,8,8a-tetrahydro-2H,7H-naphthalene1,6-dione (5a) up to 57%. To improve the reaction yield of the product, we developed the new method that the pressure and temperature manipulation in one pot by using sub- and supercritical CO2 can be achieved successfully to obtain 5a in high yield (95%) and high selectivity (95%) without any modifications of the catalysts. This simple method could be applied for various ketones and various catalysts to synthesize fused polycyclic compounds. 1. Introduction Robinson annelation1,2 of cyclohexanones, or cyclohexanediones, with methyl vinyl ketone and its homologues is an important process for the formation of cyclic structures in the synthesis of many natural products, such as the steroid series, which is a Michael reaction, followed by an aldol condensation, and then dehydration (Scheme 1). Though Robinson annelation has been widely used in organic synthesis, a longer reaction time than 6 h and several reaction steps with organic solvents such as benzene, toluene, xylene, and methanol are usually required to achieve conventional yields.3,4 To find the solution to these serious defects, one-step Robinson annelation has been developed, but the reaction had been conducted with organic solvents and the yields still remained low.5 To develop the solvent-free and simple organic reaction method, utilization of carbon dioxide (CO2), especially supercritical CO2 (scCO2), has been noted as an environmentally benign way in recent years because CO2 is considered to be a nontoxic, nonflammable, accessible, and inexpensive medium for extraction, precipitation, and reaction. Furthermore, CO2 also has several advantages favorable for chemical reaction such as a mild critical point (Tc ) 304.2 K; Pc ) 7.38 MPa) and easily tunable physicochemical properties such as solubility, density simply by manipulating the pressure and temperature, and faster mass transfer than conventional organic solvents.6,7 Recently, we successfully accelerated a reaction such as CO2 fixation by using scCO2 without catalysts8 and controlled the ratio of aldol to enal products of aldol condensation by pressure manipulation without any organic solvent in the presence of a heterogeneous MgO catalyst, which is environmentally friendly and is especially a great advantage to catalyst/product separa* To whom correspondence should be addressed. Tel.: +8122-237-5211. Fax: +81-22-237-5215. E-mail: h-kawanami@ aist.go.jp. † National Institute of Advanced Science and Technology. ‡ Hokkaido Industrial Research Institute.
tion.9-11 In this work, we investigate the new reaction method for Robinson annelation by simple pressure and temperature manipulation in sub- and supercritical CO2 in the presence of various heterogeneous catalysts to synthesize fused polycyclic compounds such as 8amethyl-3,4,8,8a-tetrahydro-2H,7H-naphthalene-1,6-dione (5a) in high yield. 2. Experimental Part 2.1. Materials. Metal oxides Al2O3, Nb2O5, La2O3, and CeO2 were purchased from Wako Pure Chemical Industries, Ltd., and MgO (JRC-MGO-4 100A), TiO2 (JRC-TIO-4), and ZrO2 (JRC-ZRO-2) were purchased from The Catalysis Society of Japan. The properties of the MgO catalyst are as follows: particle size, ca. 14 nm; Brunauer-Emmett-Teller surface area, 120 m2/ g. These metal oxides were used after calcination at 200 °C for 1 h under reduced pressure. 2.2. Experimental Methods. The reaction apparatus is shown in Figure 1. The typical experimental procedure is as follows: 2-methylcyclohexa-1,3-dione (0.13 g, 1.0 mmol), 3-buten-2-one (0.1 mL, 1.2 mmol), and MgO (50 mg) were charged into a 25-cm3 stainless steel reactor at room temperature. The reactor was heated at 100 °C, and CO2 was subsequently charged into the reactor using a high-pressure liquid pump (Jasco CO2 delivery pump) and compressed to the desired pressure. The reactions were started by stirring the mixture, which was continued for 2 h. After the first reaction, the reactor was then heated again at 180 °C within 5 min, further CO2 was introduced, and the mixture was reacted for 2 h. After the second reaction, the reactor was cooled quickly to 0 °C with ice and the pressure was released slowly. MgO was filtered off, the products were extracted by ether, and the yields were determined by GC-MS/MS (Varian CP3800/1200L) with decane as an internal reference. The isolated yields were obtained after purification by silica gel (Si60) chromatography with n-hexane + AcOEt as the eluent. The recovered MgO was reused for the next experiment after washing with ether and acetone and calcination at 200 °C under reduced pressure.
10.1021/ie050254z CCC: $30.25 © 2005 American Chemical Society Published on Web 08/31/2005
9562
Ind. Eng. Chem. Res., Vol. 44, No. 25, 2005
Scheme 1. Robinson Annelation and Hetero-Diels-Alder Reaction
Table 1. Robinson Annelation of 2a with 1a in the Presence of Metal Oxide Catalystsa run
substrate
catalyst
pressure/MPa
temp/°C
convn of 1a/%
yield of 6/%
yield of 3a/%
yield of 4a/%
yield of 5a/%
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
1a, 2a
MgO Al2O3 TiO2 ZrO2 Nb2O5 La2O3 CeO2 MgO Al2O3 TiO2 ZrO2 Nb2O5 La2O3 CeO2 MgO Al2O3 TiO2 ZrO2 Nb2O5 La2O3 CeO2
8 8 8 8 8 8 8 12 12 12 12 12 12 12 20 20 20 20 20 20 20
35 35 35 35 35 35 35 100 100 100 100 100 100 100 180 180 180 180 180 180 180
35 39 36 19 47 38 51 90 84 88 89 44 81 67 86 88 90 88 86 87 88
18 27 20 13 34 31 38 9 3 2 1 24 3 26 8 0 0 0 0 0 0
17 12 16 6 13 7 13 77 72 82 78 0 60 35 21 17 51 32 53 0 36
0 0 0 0 0 0 0 0 8 3 6 14 16 3 0 19 0 16 1 31 24
0 0 0 0 0 0 0 4 0 0 0 0 0 0 57 52 38 40 33 57 58
a
The reaction time is 2 h.
Figure 1. Batchwise reaction apparatus for Robinson annelation under sub- and supercritical CO2 in the presence of metal oxide catalysts.
3. Results and Discussion Robinson annelation of 2-methylcyclohexane-1,3-dione (2a) with 3-buten-2-one (1a) was conducted in batchwise operation under various conditions in the presence of metal oxide catalysts, and the results of the preliminary screening were shown in Table 1. The first reaction of the Michael reaction of 1a with 2a was proceeded at temperatures of 35, 100, and 180 °C, whereas the yield of 3a at 35 °C was lower than that of byproduct 6 (runs 1-4) because 1a can be dissolved well into the CO2 phase at 35 °C12 and byproduct 6, which is formed by hetero-Diels-Alder cycloaddition (Scheme 1a),13 is known to proceed favorably under supercritical and pressurized conditions.14,15 This formation of 6 interferes with the Michael reaction of 1a and 2a, leading to a decrease in the yield of 3a. At the higher temperature of 100 °C, the first reaction from 1a to 3a is proceeded to attain a
better yield of 3a than of 6. MgO, ZrO2, TiO2, and Al2O3 can effectively activate the reaction to afford good conversions of 1a (runs 5-8) at 100 °C and 12 MPa, and when only MgO was used, the desired product 5a was formed but the yield was very low. To obtain 5a with higher yield, the reaction was operated at 20 MPa and 180 °C. However, the conversion of 1a at 180 °C was almost the same as that at 100 °C and the yield of 5a was improved from 4% to 57% with the formation of 8% of 6 when MgO was used. On the other hand, in the presence of other catalysts, only 5a was obtained with good yields without forming 6 (runs 16-21), whereas the yield of 5a was still insufficient compared to that in the conventional organic solvents (74%). To obtain better yield and selectivity of 5a, we studied the details of the reaction at various pressures and at 100 and 180 °C under catalysis by MgO, which derives the best conversion of 1a at 100 °C and one of the best yields of 5a at 180 °C, and the results are shown in Figures 2 and 3. At various pressures and a constant temperature of 100 °C (Figure 2), the reaction produced 5a as one of the minor products and 2-methyl-2-(3oxobutyl)cyclohexane-1,3-dione (3a) as a main product, which is the product of the Michael reaction between 1a and 2a. The selectivities of 5a and byproduct 6 decrease with an increase in the pressure up to 10 MPa, whereas the selectivity of 3a increases, reaching a maximum selectivity of 95% at 10 MPa near the critical pressure. Especially, at the near-critical pressure of 10 MPa, the yield of product 3a attained up to a maximum of 78% with an excellent selectivity of 92%, but the yield of 5a was very low. It is considered that, in the near-
Ind. Eng. Chem. Res., Vol. 44, No. 25, 2005 9563 Table 2. Robinson Annelation of 2a with 1a in the Presence of Metal Oxide Catalystsa run
substrate
1
1a, 2a
catalyst MgO
2
MgOb
3
Al2O3
4
TiO2
5
ZrO2
6
ZrO2
7
Nb2O5
8
La2O3
9
CeO2
10
pressure/MPa
time/h
temp/°C
10 (first) 20 (second) 10 (first) 20 (second) 10 (first) 20 (second) 10 (first) 20 (second) 10 (first) 20 (second) 10 (first) 20 (second) 10 (first) 20 (second) 10 (first) 20 (second) 10 (first) 20 (second)
2 2 2 2 2 5 2 5 2 2 2 3 2 4 2 4 2 5 c
100 180 100 180 100 180 100 180 100 180 100 180 100 180 100 180 100 180 refluxd
recovery of 1a/%
yield of 6/%
yield of 3a/%
yield of 5a/%
0
2
3
95
0
2
4
94
0
1
0
99
4
0
2
94
9
4
35
38
0
1
0
97
1
0
0
99
5
0
4
91
1
1
2
94 74
a The reaction time is 2 h. b The catalyst that was used in run 1 was reused. c The pressure was atmospheric pressure. time was 4 h (methanol reflux) + 16 h (benzene reflux) under ambient conditions.
Figure 2. Pressure dependence of the selectivity of each product (3a, 5a, and 6) under scCO2 at 100 °C.
Figure 3. Pressure dependence of the selectivity of each product (3a, 5a, and 6) under scCO2 at 180 °C.
critical density (0.189 g/cm3) at 10 MPa, the local concentration of 2a with 1a would be larger than that at lower pressure,16,17 and the product 3a is less soluble in the CO2 phase as compared to 1a and 2a,12 resulting in the promotion of the Michael reaction of 2a with 1a. On the other hand, at a higher pressure than 10 MPa, aldol condensation from 3a to 5a was accelerated under supercritical conditions in the presence of the MgO catalyst, giving an increasing yield of 5a and a decreasing yield of 3a. However, byproduct 6 is also formed by hetero-Diels-Alder cycloaddition. Accordingly, the Michael reaction from 2a to 3a would be suitable for
d
The reaction
reaction conditions of 10 MPa and 100 °C, but the yield of 5a was not improved by only pressure manipulation. The reaction at a higher temperature of 180 °C was further investigated at several pressures, leading to a better yield of 5a than 3a, with a maximum yield of 57% and a selectivity of 65% at 20 MPa. The supercritical conditions at 20 MPa and 180 °C would be favorable for aldol condensation followed by dehydration to give 5a. However, both selectivities of 3a and 5a were decreased with an increase in the pressure above 20 MPa because of interference of the generation of 3a and 5a by the formation of 6. To further improve the yield of 5a in one pot, we attempted the reaction at a lower pressure of 10 MPa at 100 °C, to begin with, and then subsequently at a higher pressure of 20 MPa and a high temperature of 180 °C in the presence of MgO, Al2O3, TiO2, ZrO2, Nb2O5, La2O3, and CeO2. Consequently, as shown in Table 2, this one-pot reaction method by varying the pressure and temperature in the presence of MgO (run 1) provides a much better yield (95%) and selectivity (95%) of 5a compared to a 74% yield of 5a synthesized step by step in organic solvents (run 10). In addition, this heterogeneous catalyst can be separated easily by only filtration and can be reused (run 2). Furthermore, this reaction method can also be applied for other catalysts (runs 3-9). When ZrO2 was used, 2 h + 2 h of reaction time was not sufficient to obtain 5a in enough yield (run 5), but the longer reaction time can afford 97% of a good yield (run 6).18 Other catalysts also have the activity for the reaction to give sufficient yields (g91%) within 7 h (e2 h + 5 h; runs 3-9), whereas reaction times in organic solvents were more than 20 h (run 10). So, this suitable manipulation of only the pressure and temperature under sub- and supercritical conditions has enabled one-pot Robinson annelation, which consists of three elemental reactions of the Michael reaction, aldol condensation, and dehydration. We also applied this pressure-temperature variable method for the one-pot Robinson annelation to other cyclic ketones, such as 2-methyl-cyclopentane-1,3-dione (2b) with 1a and 2b with 1b, as shown in Scheme 2. When 2b was used, polycyclic compounds 5b and 5c were also obtained in good yields of 94% and 91%, respectively, under conditions similar to those of the MgO catalyst in the case of 2a. Therefore, acceleration
9564
Ind. Eng. Chem. Res., Vol. 44, No. 25, 2005
Scheme 2. Robinson Annelation with 1a, 1b, and 2b by Pressure and Temperature Manipulation
only of Robinson annelation in one pot to obtain polycyclic compounds in good yield and good selectivity can be achieved by simple gradual pressure and temperature manipulation, and this method can be used for other reactions. 4. Conclusion In conclusion, we investigated Robinson annelation by using scCO2 in the presence of various metal oxide catalysts to develop one-pot organic solvent-free methods. Our new method, gradual pressure and temperature manipulation under sub- and supercritical conditions with heterogeneous catalysts such as MgO, Al2O3, TiO2, ZrO2, Nb2O5, La2O3, and CeO2, can achieve onepot Robinson annelation in high yields within a shorter reaction time compared to that in conventional organic solvents. For example, we showed that by adjustment of the pressure and temperature 5a was obtained in higher yield than 94%, and various ketones can be applied for the synthesis of fused polycyclic compounds. Acknowledgment This work was supported in part by CREST from Japan Science and Technology Corp., Industrial Technology Research Grant Program in ’03 from NEDO of Japan, and Ministry of Education, Science, Sports and Culture [Grant-in-Aid for Young Scientists (B), 16760619, 2004]. Literature Cited (1) Rapson, W. S.; Robinson, R. Experiments on the Synthesis of Substances related to the Sterols. Part II. A New General Methods for the Synthesis of Substituted cyclo-Hexenones. J. Am. Chem. Soc. 1935, 57, 1285.
(2) Bergmann, E. D.; Ginsburg, D.; Pappo, R. The Michael Reaction. Org. React. 1959, 10, 179. (3) Kitahara, Y.; Yoshikoshi, A.; Oida, S. Total synthesis of dolabradiene. Tetrahedron Lett. 1964, 5, 1763. (4) Dutcher, J. S.; Macmillian, J. G.; Heathcock, C. H. Pentacyclic Triterpene Synthesis. 5. Synthesis of Optically Pure Ring AB Precursors. J. Org. Chem. 1976, 41, 2670. (5) Miyamoto, H.; Kanetaka, S.; Tanaka, K.; Yoshizawa, K.; Toyota, S.; Toda, F. Solvent-Free Robinson Annelation Reaction. Chem. Lett. 2000, 888. (6) Chemical Synthesis Using Supercritical Fluids; Jessop, P. G., Leitner, W., Eds.; Wiley-VCH: Weinheim, Germany, 1999. (7) High-Pressure Chemistry; Eldik, R., Klarner, F.-R., Eds.; Wiley-VCH: Weinheim, Germany, 2002. (8) Kawanami, H.; Ikushima, Y. Chemical fixation of carbon dioxide to styrene carbonate under supercritical conditions with DMF in the absence of any additional catalysts. Chem. Commun. 2000, 2089. (9) Matsui, K.; Kawanami, H.; Ikushima, Y.; Hayashi, H. Control of the self-aldol condensation by pressure manipulation under compressed CO2. Chem. Commun. 2003, 2502. (10) Matsui, K.; Kawanami, H.; Hayashi, H. Aldol reactions of propanal using MgO catalyst in supercritical CO2. Stud. Surf. Sci. Catal. 2004, 153, 363. (11) Kawanami, H.; Ikushima, Y. Promotion of one-pot Robinson annelation achieved by gradual pressure and temperature manipulation under supercritical conditions. Tetrahedron Lett. 2004, 45, 5147. (12) We confirmed the phase at each condition by using a highpressure view cell. The detail phase diagram is under consideration. (13) Rimmelin, J.; Jenner, G.; Abdi-Oskoui, H. Study of pericyclic reactions under pressure. V. Reactions of Diels-Alder competitives between derives carbonyls insatures. Bull. Soc. Chim. Fr. 1977, 341. (14) Jenner, G. High-pressure kinetics of Lewis acid-catalyzed cycloadditions and ene reactions. A convincing method for mechanistic delineation. New J. Chem. 1997, 21, 1085. (15) Ikushima, Y.; Saito, N.; Arai, M. Supercritical Carbon Dioxide as Reaction Medium: Examination of Its Solvents Effects in the Near-Critical Region. J. Phys. Chem. 1992, 96, 2293. (16) Ikushima, Y.; Saito, N.; Arai, M.; Blanch, H. W. Activation of a Lipase Triggered by Interactions with Supercritical Carbon Dioxide in the Near-Critical Region. J. Phys. Chem. 1995, 99, 8941. (17) Randolph, T. W.; Clark, D. S.; Blanch, H. W.; Prausnitz, J. M. Enzymatic Oxidation of Chlesterol Aggregates in Supercritical Carbon Dioxide. Science 1988, 239, 387. (18) The yield of 6 was decreased from 4% to 1% (runs 5 and 6) because retro-hetero-Diels-Alder reaction occurred.
Received for review February 27, 2005 Revised manuscript received May 26, 2005 Accepted June 27, 2005 IE050254Z
