466133 Hydroformylation and Tandem Isomerization-Hydroformylation of Long-Chain Olefins: Mechanism, Kinetics and Optimal Reaction Control

Tuesday, November 15, 2016: 4:21 PM
Franciscan A (Hilton San Francisco Union Square)
Andreas Jörke, Institute of Process Engineering, Otto von Guericke University, Magdeburg, Germany, Andreas Seidel-Morgenstern, Physical and Chemical Foundations of Process Engineering, Max Planck Institute for Dynamics of Complex Technical Systems, Magdeburg, Germany and Christof Hamel, Anhalt University of Applied Sciences, Köthen, Germany

Introduction
The Hydroformylation of n-olefins using Rh-Bisphosphite catalysts (e.g. Rh-Biphephos) for the production of terminal aldehydes is known to suffer from side reactions, such as double-bond isomerization and hydrogenation of the substrate [1]. However, the ability to isomerize the n-olefins to their thermodynamic equilibrium composition allows such catalysts to perform the transformation of complex n-olefin mixtures with internal double-bond to valuable terminal aldehydes [2]. To exploit such processes on a technical scale, it is necessary to understand the mechanism of the reactions and tandem-effects in order to find a good kinetic model for reactor design and operation with an optimal yield of terminal aldehydes. This motivated the experimental and theoretical analysis of the hydroformylation reaction network including isomerization and hydrogenation of 1-decene and iso-decenes with terminal and internal double-bond, respectively. Based upon an extended reaction mechanism, detailed kinetic models were derived including catalyst equilibria. Fitting the model to experimental data was supported by parameter subset selection to avoid overparameterization [3]. With this model, optimal reaction control strategies were generated.

Experimental
The kinetic experiments were performed with 1-decene and an equilibrium mixture of iso-decenes in a thermomorphic multicomponent solvent system consisting of DMF, dodecane and n-decenes. A Rh-Biphephos catalyst was used in all experiments with a Rh-to-olefin ratio of 1:10000-1000. All reactions were realized in a 90 ml high pressure multiple batch reactor system (Parr Instrument Co.) under careful consideration of the catalyst pretreatment at 15-18 bar syngas and a molar Rh-to-ligand ratio of 1:3. Temperature, total Syngas pressure, partial pressures of CO and H2 as well as initial substrate concentrations were varied systematically in 26 batch and semi-batch experiments. Gas chromatography (GC) was used for quantification.

Calculations
The parameter subset selection calculations and the parameter estimation were performed using Matlab 2012a with standard solvers. Optimal reaction control strategies were generated via maximizing the yield of undecanal subject to balance equations, kinetics and thermodynamics with time dependent state variables (temperature, partial pressures of CO/H2) as degrees of freedom. This optimal control problem was transformed to a large NLP using orthogonal collocation and solved in AMPL.

Results
The experimental results showed the expected high isomerization activity of the catalyst as well as the highly regioselective hydroformylation. The activity towards hydrogenation was always low. The presence of CO affected the active catalyst and inhibited the isomerization reaction drastically, whereas the hydroformylation is promoted by CO and H2. Therefore, increasing the syngas pressure lead to an increase in aldehyde yield if 1-decene was used as substrate. However, increasing syngas pressures applied to an equilibrated mixture of iso-decenes with internal double-bond counterintuitively decreased the aldehyde yield. However, the ratio of linear to branched aldehydes remained almost constant at about 97:3 in all cases. In total, 26 experiments including batch and semi-batch mode were used for parameter estimation and 14 parameters were estimated with low 95% confidence intervals (<10%). The resulting model was able to describe all reactions under all reaction conditions with low deviations (see Fig. 1).

Fig. 1: Selected model calibration results: a) Isomerization of 1-decene (Rh:Olefin = 1:10000, 0 bar CO/H2,
105 °C); b) Hydrogenation of 1/iso-decene mixture (Rh:Olefin = 1:10000, 10 bar H2, 105 °C); c) Hydroformylation of 1-decene (Rh:Olefin = 1:10000, 10 bar CO/H2, 105 °C); d) Tandem-Isomerization-Hydroformylation of an equilibrated iso-decene mixture (Rh:Olefin = 1:1000, 5 bar CO/H2, 105 °C).

The optimization results suggest to perform the reactions at low temperatures and high syngas pressures for maximum aldehyde yield it if 1-decene is used as substrate. The opposite is the case if an equilibrated mixture of iso-decenes with internal double-bond is used as substrate (see Fig. 2) which is a result of the coupling of isomerization and hydroformylation in the sense of a tandem-reaction.

Fig. 2: Selected optimal reaction control results for tandem isomerization-hydroformylation (Rh:Olefin = 1:1000): a) Concentration-time curves for maximum undecanal yield; b) Optimal temperature profile;
c) Optimal partial pressure profiles.

Acknowledgement
This work is embedded in the collaborative research center „SFB/TRR-63 InPROMPT - Integrated Chemical Processes in Liquid Multiphase Systems“ and the authors gratefully thank the German Science Foundation for financial support.

Literature
[1] R. Franke et al., Chemical Reviews 2012, 112, 5675−5732
[2] A. Behr et al., Journal of Molecular Catalysis A: Chemical 2003, 206, 179–184
[3] A. Jörke et al., Chemie Ingenieur Technik 2015, 87, 713–725


Extended Abstract: File Not Uploaded
See more of this Session: Reaction Chemistry and Engineering II
See more of this Group/Topical: Catalysis and Reaction Engineering Division