An effective and recyclable catalyst for hydrogenation of α,β-unsaturated aldehydes into saturated aldehydes in supercritical carbon dioxide

Abstract

A carbon supported palladium catalyst (Pd/C) is successfully used in selective hydrogenation of α,β-unsaturated aldehydes in scCO₂ and under solventless reaction conditions. The reactions take place very rapidly in scCO₂ and the average turnover frequency values for cinnamaldehyde and crotonaldehyde at 100% conversion maximally reached 3.1 and 11.6 s⁻¹, respectively. The catalyst can be recycled several times without loss of activity and selectivity. The products (liquid), solvent (scCO₂) and catalyst (solid) can be easily separated by a simple phase separation process. The present reaction is an ideal green chemical process in the view of industrial applications.

---

Green Context

The conversion of unsaturated aldehydes into saturated ones has industrial and biological relevance. The hydrogenation reaction requires a catalyst but even with very active palladium-based catalysts, reaction rates can be slow due to the poor solubility of the H₂ in organic solvents. By substituting supercritical carbon dioxide as solvent, this problem is overcome and we relieve the environmental problems associated with the use of VOCs. Here a novel methodology for the hydrogenation of unsaturated aldehydes based on the use of Pd/C catalyst and scCO₂ solvent is described.

JHC

Introduction

Much attention has been paid for selective hydrogenation of α,β-unsaturated aldehydes into the corresponding unsaturated alcohols owing to their applications in the synthesis of fine chemicals such as perfumes and pharmaceuticals.^1,2 The production of saturated aldehydes from unsaturated ones has also industrial and biological applications. For example, hydrocinnamaldehyde derived from cinnamaldehyde hydrogenation can be used in the synthesis of an intermediate reagent of anti-viral pharmaceuticals, particularly HIV protease inhibitors.³ Hydrogenation of α,β-unsaturated aldehydes to specific products (Scheme 1) such as saturated aldehydes and alcohols, as well as unsaturated alcohols, can be achieved by choosing suitable metal catalysts. Palladium is one of the most active and selective catalysts for the hydrogenation of unsaturated aldehydes to produce saturated aldehydes in conventional organic solvents.^4–7 However, the rate of catalytic hydrogenation in a gas–liquid system is not as high as wished due to the low solubility of gaseous H₂ in conventional organic solvents. In contrast, H₂ is completely miscible with supercritical carbon dioxide (scCO₂), and the proper use of scCO₂ in heterogeneous catalysis can eliminate gas–liquid phase transfer resistance, enhance reaction rate and mass- and heat-transfer, control selectivity, and/or improve catalyst lifetime and regeneration.^8–10 It has been reported that scCO₂ medium can improve the reaction rate and selectivity of cinnamaldehyde hydrogenation over supported platinum catalysts.^11,12 In the present work, a palladium supported on carbon catalyst (10% Pd/C) has been used for the selective hydrogenation of α,β-unsaturated aldehydes such as cinnamaldehyde and crotonaldehyde in scCO₂. The influence of some parameters such as H₂ pressure, CO₂ pressure, temperature, and reaction time has been investigated, and the effectiveness of scCO₂ reaction medium has been compared with organic solvents and solventless conditions. The catalyst recycling has also been examined in scCO₂. We report that a 10% Pd/C catalyst in scCO₂ is a highly efficient catalytic system for α,β-unsaturated aldehyde hydrogenation into saturated aldehydes.

Scheme 1 Hydrogenation of α,β-unsaturated aldehydes.

Results and discussion

Table 1 shows the results of hydrogenation of cinnamaldehyde and crotonaldehyde with Pd/C catalyst in different solvents. It is evident that only C=C double bond was hydrogenated in the hydrogenation of crotonaldehyde in scCO₂, producing butanal with 100% selectivity. However, the diacetal was formed as a byproduct in 1-propanol, which is a more active organic solvent compared with toluene. When the reaction was conducted in an apolar solvent such as toluene or under solventless conditions, similar selectivity to butanal was obtained as that obtained in scCO₂, but the reaction rate is lower. In the case of cinnamaldehyde, both C=C and C=O bonds were hydrogenated and hydrocinnamaldehyde was formed predominantly over hydrocinnamyl alcohol, but no cinnamyl alcohol was formed under the reaction conditions used. The reaction rate and selectivity to hydrocinnamaldehyde were improved in scCO₂ compared with 1-propanol, in which the diacetal was also produced as a byproduct, but the selectivity to hydrocinnamaldehyde was lower compared with that in scCO₂. When 1-propanol and scCO₂ were used at the same time, both the reaction rate and selectivity were improved compared with that obtained in 1-propanol alone. When the reaction was conducted in toluene and under solventless conditions, the selectivity to hydrocinnamaldehyde was almost the same as that obtained in scCO₂, but the conversion is lower. Similar product distributions were previously found for Pd catalyzed coupling reactions of iodoarenes in scCO₂ and solventless conditions.¹³ It can be seen from the results in Table 1 that the TOF was remarkably enhanced in scCO₂ under identical reaction conditions. Extremely high TOF values of 3.1 and 11.6 s⁻¹ were obtained for cinnamaldehyde and crotonaldehyde at 100% conversion, respectively. The amount of cinnamaldehyde dissolved in scCO₂ is estimated to be less than 3% by visual inspection under the reaction conditions, which suggests that the reactions proceeded mainly in the liquid phase even though scCO₂ was used as the reaction medium. Therefore, the same reaction mechanism can be proposed in solventless and scCO₂ conditions. The improvement of the reaction rate in scCO₂ can be explained by the fact that H₂ is completely miscible in scCO₂ and the liquid substrate dissolves a high quantity of CO₂, resulting in an ‘expanded liquid’, and thus significantly increasing the H₂ concentration in the vicinity of the solid catalyst.¹⁴ The reaction rate was shown to increase with increasing H₂ pressure.^6,11,15 This is in agreement with the results given in Table 2, showing the results of cinnamaldehyde hydrogenation under different reaction conditions. The conversion increases with increasing H₂ pressure, while the changes in the selectivity are negligible. It was reported that the liquid phase hydrogenation of cinnamaldehyde over Pd/C catalyst in propanol⁶ and Pt/SiO₂ catalyst in ethanol¹⁵ was first order to H₂ pressure. The same result was also reported in scCO₂ with Pt/Al₂O₃ catalyst.¹¹ These results are consistent with the present results at H₂ pressure up to 2 MPa.

Table 1 Results of hydrogenation of α,β-unsaturated aldehydes with 10% Pd/C catalyst

Reactant	Solvent	Time/min	Conversion (%)	TOF^a/s^−1	Selectivity (%)
Reaction conditions: catalyst 10% Pd/C 0.01 g (including Pd 0.0094 mmol) cinnamaldehyde 2.5 g (18.8 mmol), crotonaldehyde 2.5 g (35.6 mmol), temperature 323 K, H2 pressure 4.0 MPa, organic solvent toluene, 1-propanol 15 ml. HCAL hydrocinnamaldehyde, HCOL hydrocinnamyl alcohol, COL cinnamyl alcohol.a TOF: moles of substrate reacted per mole of exposed surface Pd atoms per second.b In the presence of 1-propanol, acetals were formed.
					HCAL	HCOL	COL
Cinnamaldehyde	CO2 (8.0 MPa)	60	100	3.1	87	13	0
	Toluene	60	18	0.6	88	12	0
	1-Propanol^b	60	75	2.3	71	22	0
	1-Propanol + CO2 (8 MPa)^b	60	90	2.8	80	17	0
	Solventless	60	64	1.9	87	13	0
					Butanal	Butanol	Butenol
Crotonaldehyde	CO2 (8.0 MPa)	10	78	27.1	100	0	0
	CO2 (8.0 MPa)	30	100	11.6	100	0	0
	Toluene	30	48	5.6	100	0	0
	1-Propanol^b	10	65	22.6	73	0	0
	Solventless	10	45	15.7	100	0	0

---

Table 2 Results for cinnamaldehyde hydrogenation under different reaction conditions in CO₂ with 10% Pd/C catalyst

Pressure (MPa)				Selectivity (%)
H2	CO2	Time/min	Conversion (%)	HCAL	HCOL	COL
Reaction conditions: cinnamaldehyde 2.5 g (18.8 mmol), 10% Pd/C catalyst 0.01 g (including Pd 0.0094 mmol), temperature 323 K, HCAL hydrocinnamaldehyde, HCOL hydrocinnamyl alcohol, COL cinnamyl alcohol.
1.0	8.0	60	46	89	11	0
2.0	8.0	60	79	88	12	0
4.0	8.0	60	100	87	13	0
6.0	8.0	60	100	87	13	0
4.0	0	40	54	87	13	0
4.0	7.0	40	72	88	12	0
4.0	8.0	40	74	88	12	0
4.0	14.0	40	83	88	12	0
4.0	8.0	20	46	88	12	0
4.0	8.0	40	74	88	12	0
4.0	8.0	60	100	87	13	0
4.0	8.0	180	100	87	13	0

---

As shown in Table 2, the conversion of cinnamaldehyde increases with increasing CO₂ pressure, while the selectivity changes very slightly. It is proposed that much more H₂ can be dissolved in the liquid phase with increasing CO₂ pressure, so that the rate of reaction is enhanced. It was found that the optical absorption spectrum of small gold particles measured in scCO₂ depended significantly on the pressure, suggesting a direct interaction between the gold particles and the scCO₂ medium.¹⁶ It was also speculated that the interaction between scCO₂ and metal particles on the support could vary the product selectivity. We have reported that the selectivity to cinnamyl alcohol increases with increasing CO₂ pressure when hydrogenation of cinnamaldehyde was conducted with Pt/Al₂O₃¹¹ and Pt/SiO₂¹² catalysts. However, the present work shows a different result in that the product distribution does not depend on CO₂ pressure and similar selectivity values are also obtained under the solventless conditions.

The conversion increased with reaction time and reached 100% after 60 min with the selectivity remaining unchanged. No further hydrogenation of C=O bonds was found even upon extending the reaction time after 100% conversion. This means that the C=O bond of hydrocinnamaldehyde is difficult to be hydrogenated and the hydrocinnamyl alcohol arises completely from cinnamyl alcohol. Otherwise, one would predict the selectivity to hydrocinnamaldehyde to decrease with time due to the decrease in the concentration of hydrocinnamaldehyde as a result of its hydrogenation to hydrocinnamyl alcohol. To verify this, the hydrogenation of hydrocinnamaldehyde was carried out under identical reaction conditions (hydrocinnamaldehyde 18.8 mmol, 10% Pd/C catalyst 0.01 g (including Pd 0.0094 mmol), H₂ 4.0 MPa, CO₂ 8.0 MPa, temperature 323 K, reaction time 1 hour) but no detectable amount of hydrocinnamyl alcohol was found. This indicates that all the hydrocinnamyl alcohol obtained in the reduction of cinnamaldehyde is completely produced through the formation of cinnamyl alcohol as an initial step in this reaction. This result is in contrast to previous studies of cinnamaldehyde hydrogenation with Pt/SiO₂ catalyst in ethanol,¹⁵ in which the hydrocinnamyl alcohol was confirmed to be produced from the hydrogenation of hydrocinnamaldehyde first and then from cinnamyl alcohol after the former was consumed completely. However, Mahmoud et al.¹⁷ reported that the reaction rate of the reduction of cinnamyl alcohol to hydrocinnamyl alcohol is about 30 times higher than that of the hydrogenation of cinnamaldehyde to cinnamyl alcohol with Pd/SiO₂ in toluene, and no cinnamyl alcohol can be detected in the products. This may also be the case for the present hydrogenation in scCO₂. Thus, we can assume different adsorption modes of CAL on the surface of Pt and Pd particles. The strength of adsorption is CAL > HCAL > COL on Pt, while the order is CAL > COL > HCAL on Pd.

The effect of additives on the reaction rate and selectivity has also been examined in the present work. An additive such as potassium acetate was reported to significantly enhance the formation of hydrocinnamaldehyde in the hydrogenation of cinnamaldehyde with Pd/C catalyst in propanol.⁶ In contrast, the addition of potassium acetate and potassium carbonate not only decreases hydrocinnamaldehyde formation but also lowers the reaction rate significantly in scCO₂ as shown in Table 3. The results obtained indicate that the reaction medium plays a very important role in α,β-unsaturated aldehyde hydrogenation.

Table 3 Effects of additives on the hydrogenation of cinnamaldehyde in scCO₂ at 323 K

		Selectivity (%)
Additive	Conversion (%)	HCAL	HCOL	COL
Reaction conditions: 10% Pd/C catalyst 0.01 g, cinnamaldehyde 2.5 g, H2 4.0 MPa, CO2 8.0 MPa, reaction time 40 min, KOAC, K2CO3 0.1 mmol, HCAL hydrocinnamaldehyde, HCOL hydrocinnamyl alcohol, COL cinnamyl alcohol.
KOAC	17	77	23	0
K2CO3	37	80	20	0
—	74	88	12	0

---

Catalyst recycling has been examined in scCO₂. 10% Pd/C catalyst (0.01 g) was used several times under the following conditions: cinnamaldehyde 1.0 g, H₂ 4.0 MPa, CO₂ 8.0 MPa, temperature 323 K, reaction time 60 min. After the catalyst was used three times, it still showed the same activity and selectivity, as shown in Table 4. From TEM observation we did not find any difference in the state of the Pd particles between fresh and recycled catalysts.

Table 4 Results of Pd/C catalyst recycling in the hydrogenation of cinnamaldehyde in scCO₂ at 323 K

		Selectivity (%)
Run	Conversion (%)	HCAL	HCOL	COL
Reaction conditions: 10% Pd/C catalyst 0.01 g, cinnamaldehyde 1.0 g, H2 4.0 MPa, CO2 8.0 MPa, temperature 323 K, reaction time 60 min, HCAL hydrocinnamaldehyde, HCOL hydrocinnamyl alcohol, COL cinnamyl alcohol.
1	100	85	15	0
2	95	86	14	0
3	96	87	13	0

---

Conclusions

Palladium supported on carbon catalyst (Pd/C) is a very effective, selective and recyclable catalyst for the hydrogenation of α,β-unsaturated aldehydes in scCO₂. The reaction rate increases with increasing H₂ and CO₂ pressures, but the selectivity does not change with the reaction conditions. Higher selectivity to saturated aldehydes can be achieved in scCO₂ and under solventless conditions, compared with that in 1-propanol. Enhanced TOF values of 3.1 and 11.6 s⁻¹ were obtained for cinnamaldehyde and crotonaldehyde, respectively, at 100% conversion in scCO₂. The separation of product, catalyst and solvent is very easy via simple phase separation and the catalyst can be reused for several times without loss in the activity and selectivity.

Experimental

A commercial 10% Pd on activated carbon (Pd/C) catalyst was purchased from Wako Pure Chem. Ind. The Pd/C catalyst was reduced by flowing H₂ at 473 K for 3 h and then stored in a desiccator before reaction. The total surface area (893 m² g⁻¹) was determined by the BET nitrogen adsorption method and the degree of palladium dispersion (0.18) was measured by hydrogen adsorption at room temperature. The size of palladium metal particles (∼5–10 nm) was measured on a JEM-2000EX electron microscope operated in the black field mode.

Hydrogenation reactions were carried out in a 50 ml high-pressure stainless steel reactor. Reactants and a set amount of catalyst were charged into the reactor, and the reactor was then sealed, flushed with 2.0 MPa carbon dioxide three times and heated to 323 K. After the introduction of hydrogen, liquid carbon dioxide was introduced into the reactor with a high-pressure liquid pump to the desired pressure. The reaction was conducted by stirring the mixture for a desired time. It was then cooled with an ice-water bath for about 10 min, carbon dioxide and hydrogen were then carefully vented, and the liquid reaction mixture was separated from the catalyst by filtration. The reaction mixture was analyzed with a gas chromatograph (HP 6890, HP 5 capillary column: 15 m, 0.32 mm, 0.2 μm) with flame ionization detector. We define the selectivity towards reaction products and the turnover frequency (TOF) as

where i is saturated aldehyde, unsaturated alcohol, or saturated alcohol.

References

1. P. Gallezot and D. Richard, Catal. Rev., 1998, 40, 81.
2. F. Delbecq and P. Sautet, J. Catal., 1995, 152, 217.
3. A. M. C. F. Castelijns, J. M. Hogeweg, and S. P. J. M. van Nispen, PCT Int. Appl., WO 9611898 Al, 25 April 1996, 14 pp; US 5811588, 22 September 1998, 6 pp.
4. G. R. Cairns, R. J. Cross and D. Stirling, J. Catal., 1997, 166, 89.
5. M. A. Aramendia, V. Borau, C. Jimenez, J. M. Marinas, A. Porras and F. J. Urbano, J. Catal., 1997, 172, 46.
6. L. Zhang, J. M. Winterbottom, A. P. Boye and S. Rarmahasay, J. Chem. Technol. Biotechnol., 1998, 72, 264.
7. C. Pham-Huu, N. Keller, G. Ehret, L. J. Charbonniere, R. Ziessel and M. J. Ledoux, J. Mol. Catal. A: Chem., 2001, 170, 155.
8. A. Baiker, Chem. Rev., 1999, 99, 453.
9. P. G. Jessop, T. Ikariya and R. Noyori, Chem. Rev., 1999, 99, 475.
10. R. S. Oakes, A. A. Clifford and C. M. Mayer, J. Chem. Soc., Perkin Trans. 1, 2001, 917.
11. B. M. Bhanage, Y. Ikushima, M. Shirai and M. Arai, Catal. Lett., 1999, 62, 175.
12. F. Y. Zhao, Y. Ikushima, M. Shirai, T. Ebina and M. Arai, J. Mol. Catal. A: Chem., 2002, 180, 259.
13. N. Shezad, A. A. Clifford and C. M. Rayner, Green Chem., 2002, 4, 64.
14. D. Chouchi, D. Gourgouillon, M. Courel, J. Vital and M. N. Ponte, Ind. Eng. Chem. Res., 2001, 40, 2551.
15. M. Shirai, T. Tanaka and M. Arai, J. Mol. Catal. A: Chem., 2001, 168, 99.
16. M. Arai, Y. Nishiyama and Y. Ikushima, J. Supercrit. Fluids, 1998, 13, 149.
17. S. Mahmoud, A. Hammoudeh, S. Gharaibeh and J. Melsheimer, J. Mol. Catal. A: Chem., 2002, 178, 161.

---

Footnote

† Japan Society for the Promotion of Science, Domestic Research Fellow.

---