Heterogeneous catalysis with continuous flow microreactors

Abstract

Packed-bed microreactors are employed under flow conditions for studies of heterogeneous catalysis: oxidation of 4-isopropylbenzaldehyde and hydrogenation of 2-methylfuran. They have been demonstrated to be a valuable platform for rapid screening of catalytic materials, efficient optimization of reaction conditions, inline monitoring of reaction progress, and extraction of kinetic parameters.

---

Introduction

Heterogeneous catalysis is central to numerous transformation processes in a wide range of applications, from the large-scale manufacturing of commodities and energy processing to environmental protection and green chemical technologies. For example, heterogeneous catalytic technologies play a crucial role in the petroleum refining industry to provide fuels and chemical feedstocks. In the automobile industry, catalytic converters are widely used to reduce the pollution from vehicle emissions. Heterogeneous catalysis is also essential for solving many of the challenging problems underlying the generation of sustainable fuels utilizing alternative feedstocks and the development of atom- and energy-efficient chemical transformations.

The design of highly selective, energy efficient, and environmentally benign catalytic processes requires detailed knowledge of the reaction kinetics, efficient evaluation of catalytic materials, and rapid optimization of reaction conditions. Advances in continuous flow microreactor technologies have brought opportunities to address these challenges.^1–4 The term “microreactor” has been traditionally used to refer to millimeter-diameter tubular reactors used in heterogeneous catalysis. It is now used to designate miniaturized reaction devices with feature sizes in the submillimeter range.⁵ Microreactors offer a number of advantages over classical catalysis setups.^5–7 For instance, heat transfer limitations often encountered in traditional large scale setups diminish the ability to study fast and highly exothermic reactions. The sub-millimeter dimensions and high surface area to volumes of microreactor devices ensure fast heat transfer allowing for precise and rapid control of reaction temperatures. Similarly, mass transfer is enhanced so that reactions can be explored in the absence of, or at least with quantifiable, mass transfer. Another challenge of macroscopic laboratory scale setups is the extensive safety precautions required to investigate high-pressure heterogeneous catalytic reactions.^8,9 The small scale of microreactors significantly reduces the needed safety infrastructure enabling expanded window of operation.^8,10

Microreactors enable precise control of reaction parameters (e.g. temperature, pressure),^11,12 and the reaction conditions can be continuously varied and monitored in flow systems during catalytic processes, greatly accelerating reaction kinetic studies and process optimization.^13,14 Moreover, with their small sizes, microreactors use small amounts of reagents (e.g. starting materials, catalysts) to perform catalytic reactions, which makes them cost-effective and suitable as high-throughput screening platforms with parallel reactions in multiple channels.^15–17 Inline chemical reaction monitoring and product analysis have been incorporated into continuous flow systems to improve the process control with microreactors.^8,18–22

Additionally, the small footprint of these miniaturized devices makes it possible to work in a safer environment by enabling safer surroundings for chemical reactions that use hazardous reactants or intermediates, isolating air-sensitive reagents, and generating less hazardous waste.^23–27

Herein we present two case studies of heterogeneous catalytic reactions in continuous flow microreactors, (1) oxidation of 4-isopropylbenzaldehyde and (2) hydrogenation of 2-methylfuran. These examples serve to illustrate the ability to handle reactive, pressurized gases, efficiently optimize reaction conditions, and extract rate and mass transfer information.

Experimental section

Studies on the aerobic oxidation of 4-isopropylbenzaldehyde (IBA) were performed in a continuous flow system with a single-channel (27 × 2 × 0.6 mm³) silicon–Pyrex microreactor (Fig. 1) that could be heated and pressurized. An array of pillars was fabricated downstream of the reactor with 25 μm intervals as a weir to hold all the materials inside the channel.^1,28 A thin layer of glass beads (75 μm) was first packed on top of the weir structure. The slurry of a catalytic material in water was then directly loaded into the channel through the inlet port using a syringe to form a tightly packed catalyst bed. The IBA solution (1.5 M in n-butyl acetate, 1–20 μL min⁻¹) was delivered into the reactor by a syringe pump and gaseous oxygen (200–900 μL min⁻¹) by a mass flow controller. The temperature of the reactor was monitored using a K-type thermocouple that was inserted into a small hole on the backside of the packaging chuck and was controlled by a PID controller (Omega, CN7833) that was connected to the heating cartridge inserted into the chuck. A 6-way valve was used to collect product samples for chemical analysis using a gas chromatograph (Agilent 6890). The size of the sampling loop is 100 μL. A stainless steel vessel was used to provide back pressure (2–6 atm) from a high-pressure nitrogen gas cylinder and to collect waste materials. The pressure drop in our reactor system is in the range of 0.7–1 atm. The time to achieve steady state appears to be dependent, to a great extent, on the liquid flow rate: lower liquid flow rates lead to longer time. For a liquid flow rate of 6 μL min⁻¹, it takes about 30 min for the system to achieve steady state. We studied a number of Pt- and Pd-based supported catalytic materials and identified Pt/Al₂O₃-5wt% (Alfar Aesar) to be the most efficient catalyst for the selective oxidation of IBA.

Fig. 1 (A) Packaged packed-bed microreactor for catalytic oxidation of 4-isopropylbenzaldehyde. (B) Front view of the reactor channel that is loaded with 11 mg of Pt/Al₂O₃-5wt% catalyst.

The experimental setup for hydrogenation reactions consisted of several high-pressure syringe pumps (Teledyne Isco 100DM) for liquid and gas delivery, a back pressure controller (Bronkhorst EL-PRESS), a temperature controller, and a microreactor. The system is operable up to 100 atm from 25 °C to 350 °C. The microreactor is made of silicon and Pyrex (Fig. 2). There are two zones in the reactor, heated and cooled, separated by a halo etched region. A halo etch removes a large cross section of silicon and forms a boundary between the two zones, which allows a temperature gradient to be sustained across the remaining silicon with a lower overall heat flux. A single gas inlet is designed and connected to each of the four main reaction channels. The width and the depth of the catalyst channel are 0.400 mm while the length of the heated region of the catalyst bed is 47 mm.⁹

Fig. 2 Schematic of the packed-bed microreactor for catalytic hydrogenation of 2-methylfuran (2MF).

The reaction zone temperature is controlled by using a temperature controller (J-Kem, Gemini-K) and a heating chuck which is made of aluminum. The chuck has two holes into which two resistive cartridge heaters (Omega, CIR-1021/120 V) are inserted. A K-type thermocouple (Omega, KMQSS-062G) is placed between the reactor and the aluminum chuck. The Gemini temperature controller provided feedback control to the cartridge heaters. Using this heating system, a reactor temperature of 25–350 °C was achieved.

In addition, the system incorporated a microfluidic device with an attenuated total reflectance (ATR) section for inline Fourier transform infrared spectroscopy (FTIR). This device is placed at the down stream of the packed bed microreactor and they are connected with a single line of 500 mm long, 0.252 mm ID PEEK tubing. A schematic of the microfluidic cell is given in Fig. 3 and the full details will be reported later.²⁹ The IR data were collected with a Bruker Vertex 70 spectrometer using a MCT (HgCdTe) detector. For each spectrum, 32 scans were added with a resolution of 4 cm⁻¹. The received catalyst powder (Alfa Aesar) was sieved and the 36–53 μm fraction was used. Before the experiments, the catalyst was reduced in the microreactor under hydrogen flow at 250 °C for 4 h with a ramp of 4 °C min⁻¹. The amount of the catalyst used in an experiment was 6 mg. In a typical experiment, neat 2-methylfuran (2MF) and hydrogen gas were delivered into the microreactor by the syringe pumps. The flow rate of 2MF was 5 μL min⁻¹ and that of hydrogen was 100 μL min⁻¹. The pressure of the system was kept at 45 atm. We investigated several Pt- and Pd-based supported catalysts and have found Pd/C-5wt% (Alfa Aesar) to be the most active and selective for our reactions.

Fig. 3 Schematic of the microfluidic flow cell with an attenuated total reflectance (ATR) section. It is composed of Pyrex and silicon. The silicon part has optical windows for infrared coupling with the device. The internal volume of the ATR part is 2 μL.

Results and discussion

Oxidation of aldehyde using oxygen or air

Selective oxidative transformations have found a wide variety of applications both for large-scale production of commodity chemicals and for synthesis of pharmaceutical compounds and fine chemicals. Increasing environmental and energy concerns are stimulating the development of greener technologies, such as to utilize oxidants that are cheap and clean, use solvents that show low environmental impact, and target reaction conditions that are mild and safe. From the standpoint of the choice of oxidizing reagents, molecular oxygen (O₂) is the ideal candidate mainly due to its abundance and ease of separation. One of the challenges is the control of selectivity to prevent secondary oxidation.³⁰ Continuous flow microreactors allow for precise control of temperature, pressure, and ratio of oxygen to other reagents, therefore provide opportunities for selectivity control. In addition, the ability of continuous scanning of reaction parameters renders efficient optimization of reaction conditions and rapid screening of catalytic materials.

We chose the oxidation of 4-isopropylbenzaldehyde (IBA) as a probe reaction for studies with microreactors because its corresponding carboxylic acid, cumic acid, is an important intermediate species in the production of pharmaceutical compounds (Scheme 1).³¹

Scheme 1 Oxidation of 4-isopropylbenzaldehyde (IBA) to form cumic acid using molecular oxygen as the oxidant.

Yield and selectivity were explored as a function of temperature and flow rates. As shown in Fig. 4A, the yield of cumic acid increases from a reaction temperature of 60 to 80 °C. A plateau is reached above 90 °C where the yield is over 95%. Therefore, 90 °C was chosen as a reaction temperature for the studies of the dependence of yield/selectivity on the molar ratio of oxygen to IBA (Fig. 4B). The experiments were carried out by varying the flow rate of the liquid phase under a constant flow rate of oxygen (300 μL min⁻¹). When the ratio increases from 0.45 to 1.1, the yield rapidly changes from 67 to 94%. Further increasing the ratio does not result in appreciable change in the yield of cumic acid. Mass balance was maintained in all experiments, indicating that no gas-phase products such as CO₂ were formed. The high selectivity (≥98%) could be associated with the efficient heat transfer and precise control of flow rates for reactants, enabled by microreactor technologies, to prevent generation of hot spots and/or secondary oxidation to produce, for example, combustion products.

Fig. 4 (A) Yield and selectivity as a function of reaction temperature. The flow rates of 4-isopropylbenzaldehyde (IBA) and oxygen were 6 μL min⁻¹ and 300 μL min⁻¹, respectively. A back pressure of 2 atm was applied. (B) Yield and selectivity as a function of molar ratio of oxygen to IBA. The reaction temperature was 90 °C and the back pressure was 2 atm. The concentration of IBA was 1.5 M in n-butyl acetate for all studies.

Air also proved to be an efficient source of oxidant for the oxidation of IBA. As shown in Table 1, when oxygen (entry 1) is replaced by air (entry 2) to maintain the same molar ratio of O₂/IBA the yield decreases from 95 to 76%. The lower yield is probably a result of shorter contact time of the reagents with the catalyst. This assertion is supported by the higher yield observed using a slightly lower flow rate of air (entry 3). Further lowering of the liquid flow rate (entries 4 and 5) leads to an increase in yield. When the liquid flow rate is 1 μL min⁻¹, the yield obtained is comparable with that using oxygen.

Table 1 Oxidation of 4-isopropylbenzaldehyde using oxygen or air^a

Entry	Oxidation reagent	Gas flow rate (μL min^−1)	Liquid flow rate (μL min^−1)	Space velocity (min^−1)	O2/IBA ratio	Yield (%)	Sel. (%)
a The concentration of 4-isopropylbenzaldehyde was 1.5 M in n-butyl acetate. The reaction temperature was 90 °C and the back pressure was 2 atm.
1	Oxygen	300	6	11	1.5	95	>99
2	Air	1500	6	56	1.5	76	>99
3	Air	1200	6	45	1.2	88	>99
4	Air	1200	2	44	3.6	90	>99
5	Air	1200	1	44	7.2	95	99

---

Importantly, the oxidation reaction appears to be much faster in microreactors than in semi-batches. When oxygen is bubbled through the liquid phase with catalyst dispersed in it in a semi-batch operation, at least a few hours are required for the reaction to complete;³¹ while the residence time for our system is estimated to be a few seconds based on the space velocity. Notably, the selectivity obtained using microreactors is higher than that with semi-batch operation,³¹ demonstrating the benefits of microreactors as a result of precise control of reaction conditions, including temperature, pressure, and flow rates of reactants/reagents.

Hydrogenation of olefin using hydrogen

Hydrogenation reactions are used extensively in petroleum refining, chemical manufacturing and the pharmaceutical industry.^32–34 Bio-oil from pyrolysis of cellulosic materials has multiple compounds with unstable functional groups which makes it susceptible to air oxidation and degassing. In addition, pyrolysis oil has a tendency to polymerize. Hydrogenation technologies make it possible to blend these biofuels with traditional fuels.³⁵

Hydrogenation reactions are challenging to run with traditional systems. The low solubility of hydrogen gas in most of the solvents leads to mass transfer limitations. To offset this difficulty, high pressure operations (up to 100 atm) are often employed, which increases the explosion risks associated with working with hydrogen. The use of small-size microreactors offers a safer, alternative approach to high-pressure hydrogenation reactions.

As an example, we investigated hydrogenation of 2-methylfuran (2MF), a model bio-oil compound, to 2-methyltetrahydrofuran (MTHF) which has been approved as an oxygenated gasoline additive in P series fuels (Scheme 2).³⁵

Scheme 2 Hydrogenation of 2-methylfuran (2MF) to 2-methyltetrahydrofuran (MTHF) using a Pd/C-5wt% catalyst.

With the help of the inline attenuated total reflectance (ATR) unit at the downstream of the microreactor, we were able to monitor the reactor effluents with FTIR spectroscopy. During the experiments, the temperature was increased from 25 °C to 160 °C with 15–20 °C intervals while the flow rates for hydrogen and 2MF were kept constant. As shown in Fig. 5, the IR spectrum of the starting material 2MF shows signals at 2930 and 3116 cm⁻¹, corresponding to ν_as(−CH₃) and ν(=C–H) vibrational modes, respectively. With the increase in temperature, two peaks emerge and grow in intensity at 2973 and 2861 cm⁻¹, assigned to the stretching bands of saturated (–C–H).³⁶ This observation is consistent with the conversion of 2MF to MTHF. Meanwhile, the intensity of the band at 3116 cm⁻¹ decreases, as expected with increased conversion with temperature. The FTIR spectra reveal no other products indicating that 2MF is selectively hydrogenated to MTHF under the reaction conditions.

Fig. 5 In-line FTIR spectra of hydrogenation reaction for 2-methylfuran at 45 atm from 25 °C to 160 °C.

Fig. 6A shows the conversion of 2-methylfuran with increasing temperature. At low conversion, the reactor can be considered to be a differential reactor, and a semilog plot of the IR absorbance at 2973 cm⁻¹versus the reciprocal temperature then yields an apparent activation energy of 10.3 kcal mol⁻¹ (inset in Fig. 6B).

Fig. 6 (A) Conversion of 2-methylfuran and (B) Arrhenius plot for low conversion data.

Conclusions

Microfabricated packed-bed reactors have been used as a laboratory tool for heterogeneous catalytic studies of multiphasic (gas-liquid-solid) reaction systems. Through the oxidation of aldehyde with oxygen/air, we have demonstrated microreactor-based optimization of reaction conditions and screening of catalytic materials. The high selectivity obtained toward the formation of carboxylic acid illustrates the advantages of microreactors to afford precise control of reaction conditions, including temperature, pressure, and molar ratio of reactants. In the hydrogenation reaction, an attenuated total reflectance unit was integrated with the reactor system to provide inline chemical analysis. A quantitative analysis of the kinetics in the microreactor indicates that the small reactor geometry and high pressure applied eliminate the mass transfer effects. The observations demonstrate the utility of microreactors as platforms for exploring reaction kinetics. Further advances could be realized by incorporating additional analytical tools such Raman spectroscopy³⁷ and temperature programmed desorption/reaction (TPD/TPR) analysis.³⁸

Acknowledgements

The authors thank the support of this work from Novartis-MIT Center for Continuous Manufacturing and by BP through the MIT Energy Initiative.

Notes and references

1. K. D. Nagy and K. F. Jensen, Chem. Today, 2011, 29, 18–21.
2. C. G. Frost and L. Mutton, Green Chem., 2010, 12, 1687–1703.
3. C. de Bellefon, N. Tanchoux, S. Caravieilhes, P. Grenouillet and V. Hessel, Angew. Chem., Int. Ed., 2000, 39, 3442–3445.
4. R. L. Hartman, J. P. McMullen and K. F. Jensen, Angew. Chem., Int. Ed., 2011, 50, 7502–7519.
5. K. F. Jensen, Chem. Eng. Sci., 2001, 56, 293–303.
6. K. Jähnisch, V. Hessel, H. Löwe and M. Baerns, Angew. Chem., Int. Ed., 2004, 43, 406–446.
7. C. Wiles and P. Watts, Chem. Commun., 2011, 47, 6512–6535.
8. J. Keybl and K. F. Jensen, Ind. Eng. Chem. Res., 2011, 50, 11013–11022.
9. J. Keybl, PhD thesis, Massachusetts Institute of Technology, 2011.
10. S. Marre, A. Adamo, S. Basak, C. Aymonier and K. F. Jensen, Ind. Eng. Chem. Res., 2010, 49, 11310–11320.
11. D. M. Roberge, B. Zimmermann, F. Rainone, M. Gottsponer, M. Eyholzer and N. Kockmann, Org. Process Res. Dev., 2008, 12, 905–910.
12. N. Kockmann, M. Gottsponer, B. Zimmermann and D. M. Roberge, Chem.–Eur. J., 2008, 14, 7470–7477.
13. J. P. McMullen and K. F. Jensen, Annu. Rev. Anal. Chem., 2010, 3, 19–42.
14. N. Zaborenko, M. W. Bedore, T. F. Jamison and K. F. Jensen, Org. Process Res. Dev., 2011, 15, 131–139.
15. B. Jandeleit, D. J. Schaefer, T. S. Powers, H. W. Turner and W. H. Weinberg, Angew. Chem., Int. Ed., 1999, 38, 2495–2532.
16. S. Senkan, K. Krantz, S. Ozturk, V. Zengin and I. Onal, Angew. Chem., Int. Ed., 1999, 38, 2794–2799.
17. W. H. Weinberg, B. Jandeleit, K. Self and H. Turner, Curr. Opin. Solid State Mater. Sci., 1998, 3, 104–110.
18. H. Fang, Q. Xiao, F. H. Wu, P. E. Floreancig and S. G. Weber, J. Org. Chem., 2010, 75, 5619–5626.
19. P. Koos, U. Gross, A. Polyzos, M. O'Brien, I. R. Baxendale and S. V. Ley, Org. Biomol. Chem., 2011, 9, 6903–6908.
20. C. F. Carter, H. Lange, S. V. Ley, I. R. Baxendale, B. Wittkamp, J. G. Goode and N. L. Gaunt, Org. Process Res. Dev., 2010, 14, 393–404.
21. W. Ferstle, T. Klahn, W. Schweikert, G. Billeb, M. Schwarzer and S. Loebbecke, Chem. Eng. Technol., 2007, 30, 370–378.
22. S. Krishnadasan, R. J. C. Brown, A. J. deMello and J. C. deMello, Lab Chip, 2007, 7, 1434–1441.
23. D. Webb and T. F. Jamison, Chem. Sci., 2010, 1, 675–680.
24. B. P. Mason, K. E. Price, J. L. Steinbacher, A. R. Bogdan and D. T. McQuade, Chem. Rev., 2007, 107, 2300–2318.
25. I. R. Baxendale, S. V. Ley, A. C. Mansfield and C. D. Smith, Angew. Chem., Int. Ed., 2009, 48, 4017–4021.
26. C. Wiles and P. Watts, Green Chem., 2012, 14, 38–54.
27. S. Saaby, K. R. Knudsen, M. Ladlow and S. V. Ley, Chem. Commun., 2005, 2909–2911.
28. K. D. Nagy, PhD thesis, Massachusetts Institute of Technology, 2011.
29. B. Ünal and K. F. Jensen, in preparation.
30. S. T. Oyama, A. N. Desikan and J. W. Hightower, in Catalytic Selective Oxidation, ed. S. T. Oyama and J. W. Hightower, 1993, pp. 1–14.
31. R. Shekhar and M. J. Girgis, U.S. Patent 6740776, Novartis AG, 2004.
32. C. H. Bartholomew and R. J. Farrauto, Fundamentals of Industrial Catalytic Processes, Wiley-AIChE, 2005.
33. J. W. Veldsink, M. J. Bouma, N. H. Schoon and A. Beenackers, Catal. Rev. Sci. Eng., 1997, 39, 253–318.
34. F. Roessler, Chimia, 1996, 50, 106–109.
35. G. W. Huber, S. Iborra and A. Corma, Chem. Rev., 2006, 106, 4044–4098.
36. Y.-T. Cheng and G. W. Huber, ACS Catal., 2011, 1, 611–628.
37. A. J. deMello, Nature, 2006, 442, 394–402.
38. U. J. Quaade, S. Jensen and O. Hansen, Rev. Sci. Instrum., 2004, 75, 3345–3347.

---

Footnote

† These authors contributed equally to this work.

---