OPTIMIZING REACTION RATES OF EXOTHERMIC EQUILIBRIUM REACTIONS BY FORCED TEMPERATURE PROFILES IN MICROREACTORS
Peter Pfeifer*, Katja Haas-Santo, Paolo Piermartini, Tina Zscherpe, Roland Dittmeyer Karlsruhe Institute of Technology, Institute for Micro Process Engineering, D-76344 Eggenstein-Leopoldshafen, Germany
Summary Technical solutions for highly exothermic, equilibrium-restricted solid-catalyzed reactions currently comprise inter-stage cooling between adiabatic sections, high dilution of reactants and/or catalyst in fixed bed reactors, or use of intensively cooled fluidized beds. Significant process intensification may be obtained by effective countercurrent cooling in microreactors. Applying two cases, the SO2 oxidation and the water gas shift reaction, this contribution analyzes the opportunities and challenges. Keywords Process intensification, Micro-reactors, Novel reactor technologies.
Introduction Microreactors have gained attention over the last two decades with respect to process intensification due to their enhanced mass and heat transfer characteristics and as well due to their low inventory of reactants when reaction rates are increased. By far now, most of the studies refer to just keeping the reactor temperature constant and thus enabling isothermal operation conditions. For exothermic equilibrium reactions this is, however, not the optimum temperature profile. For these reactions with reaction order greater that zero, the initial rate is mainly influenced by the Arrhenius dependence and thus high reaction temperatures are favorable. With increasing conversions the approach to equilibrium, however, will reduce the rate considerably and a lower reaction temperature should be employed. The maximum rate r as a function of the conversion X can be found by letting
dr ( X ) = 0. dT
By using counter-current cooling in microreactors the optimum falling temperature profile may be approximated in an effective way. Laboratory microreactors are usually small in size, i.e. channel length is below 10 cm, so heat conduction in the microstructure may significantly contribute to the heat flux along the reaction zone. If this axial heat flux is within the range of the heat to be transferred, the desired temperature gradients may be significantly reduced. On the contrary, the conductive heat flux in the microstructure can be used on purpose to establish a certain temperature profile, but due to the fact that the heat production rate is usually decreasing with
increasing conversion, heat removal by effective cooling may be higher than desired at the reactor outlet.
Methods For our study we selected SO2 oxidation and water gas shift as test reactions. In both cases we investigated the reaction kinetics mainly under isothermal conditions in conventional laboratory and stacked microstructure foil reactors. Details about the reactors can be found in previous publications1,2. Catalysts applied for SO2 oxidation and water gas shift reaction are Pt on TiO2 and Pt/CeO2 on Al2O3, respectively. Both were prepared by sol-gel synthesis of the support3,4 on preannealed foils (Titanium and Stainless Steel, respectively) followed by wet impregnation leading to Platinum and Cerium oxide. In case of SO2 oxidation the flow rate of SO2 was 10-100 ml(STP)/min with a stoichiometry SO2/O2 of 0.1-1 and 0.1-0.5 s mean residence time at 250 mg of catalyst. For the water gas shift reaction, experiments were conducted at 1-45 bars at 500-1500 ml(STP)/min flow between 300 and 550°C and with 200 mg of catalyst. Modeling of the reactions was done with gPROMS 2.3 using a 1-D reactor model. Mass transfer and axial dispersion were neglected based on initial considerations. For a reactor design with counter-current cooling, the heat transfer between the reaction zone and the heat conduction in the reactor material were considered in 2-D by applying a semi-porous body model.
* To whom all correspondence should be addressed:
[email protected]
250
power /W
200
150
100
80
900
850
Temperature / K
For SO2 oxidation we used a model developed in cooperation with Fila et al.5 to initially judge the heat flux to be transferred from the reaction zone to the cooling channels (see Fig. 1). It clearly demonstrates that too effective cooling in countercurrent mode could lead to quenching of the reaction - starting from the reactor outlet.
our experimental results. The model suggests that the expected conversion for a linear decrease of temperature is above the conversion for isothermal operation both for low and high operating temperature.
Conversion (decreasing T)
60
800
Conversion T = 873K 750
40
Temperature
700
20 650 0
Conversion T = 623K
600
CO conversion / %
Results
550 0.00
50
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
Reactor length / m 0 0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
reactorlenght / m
Figure 1: Calculated differential cooling power needed to establish the optimal temperature profile as a function of the reactor length for SO2 oxidation. To avoid such quenching, a more detailed investigation of the importance of the axial heat flux along the reactor is needed. Investigating a stack of multiple microstructured foils for the reaction and only one cooling foil, the result was that the effect of quenching can be avoided. The semiporous body model gave us an indication that an increase of the foil ratio (reaction vs. cooling) increases the axial temperature gradients, while the temperature of the stack of reaction foils in such an arrangement almost remains constant at every axial position in the direction of the heat transfer (see Fig. 2).
For the conference we will report additional results based on kinetic rate expressions adapted to experimental data and a more detailed study on the counter-current reactor design for both reactions.
Conclusion Microreactors may not only offer advantages in terms of heat and mass transfer under isothermal operation, but can also be used to establish a certain temperature profile in order to enhance the conversion of exothermic equilibrium-restricted reactions. However, the design of the reactor for counter-current cooling mode is challenging and needs to be investigated by a sufficiently detailed simulation approach.
References
880 860
Temperature / K
Figure 3: Calculated conversion for isothermal operation at low and high temperature and for a linear falling temperature profile for the water gas shift reaction.
840 820 800 780
Re 0.0015 ac tor 0.0010 he igh 0.0005 t/ m 0.0000
0.00 0.02 0.04 0.06
/ gth len r o act Re
m
0.08
0.10
Figure 2: Calculated temperature profile in SO2 oxidation inside a stack of 200 x 200 µm channels in 300 µm thick foils for reaction and only one for cooling; temperature as a function of the reactor length and, rectangular, in the direction of heat transfer (reactor height). In the water gas shift reaction, so far, we used a kinetic rate expression for a copper catalyst system6 with the preexponential factor adjusted to obtain conversions close to
(1) Pfeifer, P; Haas-Santo, K.; Thormann, J.; Schubert, K.; One pass synthesis of pure sulphur trioxide in microreactors, Chimica Oggi 2007, 2, 42-46. (2) Piermartini, P.; Pfeifer, P.; Schaub, G.; Water Gas Shift Reaction in Microstructured Reactors, submitted to 9th NGCS, Lyon, France, May 30th - June 3rd, 2010 (3) Oezer, N.; Cronin, J. P., Akyuz, S.; Electrochromic Performance of Sol-Gel deposited CeO2 Films, SPIE 1999, 3788, 103 (4) Haas-Santo, K.; Fichtner, M.; Schubert, K; Preparation of microstructure compatible supports by sol-gel synthesis for catalyst coating, Appl. Cat. A 2001, 220, 79 (5) K. Haas-Santo, K; Kraut, M; Pfeifer, P.; Schubert, K; Bernauer, M.; Fila, V.; Sulfur dioxide oxidation reactor with internal heat profile for a one-step synthesis of sulfur trioxide, 18th CHISA, Praha, CZ, August 24-28, 2008 (6) Van den Bussche, K. M.; Froment, G.F.; A steady-state kinetic model for methanol synthesis and the water gas shift reaction on a commercial Cu/ZnO/Al2O3 catalyst, J. Catal., 1996, 161, 1
