High-pressure gas solubility in multicomponent solvent systems for hydroformylation. Part II: Syngas solubility

https://doi.org/10.1016/j.supflu.2014.01.017Get rights and content

Highlights

  • New experimental data of syngas solubility in hydroformylation solvents.

  • Increasing syngas solubility with increasing temperature modeled by PC-SAFT using temperature-independent parameters.

  • Qualitative predictions of multicomponent systems containing syngas using PC-SAFT.

Abstract

High-pressure solubility of syngas with a molar ratio of hydrogen (H2) and carbon monoxide (CO) of 1:1 was investigated in various solvents like n-decane, dimethylformamide (DMF), 1-dodecene and n-dodecanal as well as in mixtures of n-decane and DMF and in a mixture of 1-dodecene, n-dodecanal, n-decane and DMF at temperatures between 302 K and 367 K and at pressures of up to 14 MPa. Moreover, the H2 solubility in 1-dodecene and n-dodecanal was measured in the same pressure and temperature range. The solubility measurements were performed in a high-pressure volume-variable view cell using a visual synthetic method. For modeling and prediction of the gas solubility (H2, CO, and syngas (H2/CO)) in the considered solvents, the Perturbed Chain Statistical Associating Fluid Theory (PC-SAFT) was used. The systems containing one gas (H2 or CO) and one solvent were modeled accurately by applying temperature-independent binary interaction parameters (kij's). These kij's were used to predict the syngas solubility in pure solvents and their mixtures without further adjustments. The kij between H2 and CO was always set to zero. The results showed that PC-SAFT is able to predict the syngas solubility in pure solvents with an average relative deviation of 3.1–12.0%. Syngas solubility in the n-decane/DMF mixture was predicted with a deviation of 7.2%.

Introduction

Syngas (CO/H2) is one of the key components in the hydroformylation process where it reacts with olefins to form aldehydes. Within this process, the solubility of syngas is of high interest for kinetic studies and process control issues as it has a significant influence on reaction rate and product composition [1]. A recent challenge in academia and industry is the homogeneously catalyzed hydroformylation of long-chain olefins since classical production approaches, such as the Ruhrchemie/Rhône-Poulenc process, fail for their processing. One promising approach for the hydroformylation of long-chain olefins to aldehydes is the application of thermomorphic multicomponent solvent (TMS) systems, which consist of at least one polar and one nonpolar solvent [2]. Within this approach, the reaction is carried out at high temperature in a single homogeneous liquid phase. After the reaction, the system is cooled down to a temperature where two liquid phases are formed, which can easily be separated. Schäfer et al. [3] introduced a possible solvent system, which has the desired attributes for TMS behavior and includes an efficient reaction performance and effective catalyst recycling for the hydroformylation of 1-dodecene. Dimethylformamide (DMF) was chosen as the polar solvent and n-decane as the nonpolar one. Next to the solvents of the TMS system also the hydroformylation reactant (olefin) and product (respective aldehyde) are present in the liquid mixture. Due to the hydroformylation reaction, their concentrations are changing over time. Finally, the syngas solubility in this quite complex mixture determines the quantity of the syngas which is available for the hydroformylation reaction. Due to the complexity of this system, it is highly desirable to have a modeling tool which is based on few experimental data and able to predict the syngas solubility at a whole variety of process conditions. The Perturbed-Chain Statistical Associating Fluid Theory (PC-SAFT) [4] can be used for that purpose. Its advantage is the remarkable accuracy in the prediction of multicomponent systems over a wide pressure and temperature range based only on the pure-component parameters and on the binary interaction parameters (kij) of the binary systems. In part I of this publication [5] we provided the required experimental data and the PC-SAFT modeling for solubility data of CO in n-decane, DMF, 1-dodecene and n-dodecanal (experimental data of CO solubility in n-decane can be found in [15]).

In this part we measured (if not available in literature) and modeled H2 solubility in n-decane, DMF, 1-dodecene and n-dodecanal to establish a basis of the binary interaction parameters needed for the prediction of syngas solubility. Pure-component parameter sets for H2, already available in literature, were compared and appropriate parameters chosen. Based on the complete set of parameters, the influence of temperature, pressure, as well as of kind and composition of the solvent system on the syngas solubility was predicted using PC-SAFT.

To validate these results, the syngas solubility in pure n-decane, DMF, 1-dodecene and n-dodecanal as well as in four mixtures of n-decane and DMF and in a mixture consisting of four liquids (1-dodecene, n-dodecanal, n-decane and DMF) were measured at conditions relevant for the hydroformylation (302–367 K and pressures up to 14 MPa) and compared to the predictions. The experimental data were gained in a high-pressure variable-volume view cell applying the visual synthetic method.

The results allow the reduction of experimental effort and enable a further specification of the process window and optimization of the reaction performance for hydroformylation reactions in TMS systems.

Section snippets

Materials

Syngas was purchased as a molar ratio of 1:1 of carbon monoxide (purity  98.0%) and H2 (purity  99.99%) from Messer Group GmbH. All gaseous and liquid components in this work were used without further purification as purchased from suppliers listed in Table 1. Analysis via gas chromatographic did not indicate any considerable impurities.

Gas-solubility measurements

The apparatus used for the gas-solubility measurements was the same as described in [5] (see Fig. 1).

We also used the same gas solubility measurement procedure as

PC-SAFT

The PC-SAFT model is described in detail in [4]. Within PC-SAFT the residual Helmholtz energy is calculated as a sum of contributions accounting for different interactions between molecules. These interactions account for the contributions of the repulsion (hard chain –hc), van der Waals attraction (dispersion – disp), association (assoc) and dipolar interactions (dipole). The residual (res) Helmholtz energy of a system is then calculated as the sum of the different Helmholtz-energy

PC-SAFT parameters

In order to predict the syngas solubility in pure solvents or in mixtures using PC-SAFT, the pure-component parameters, as well as the binary interaction parameters kij for each pair of components are required.

Conclusion

Within this work syngas (H2/CO) solubility in TMS systems used for the hydroformylation of long-chain olefins was investigated experimentally and through modeling using PC-SAFT.

The measurements were performed in a high-pressure variable-volume view cell by using a synthetic method (cloud-point measurements). Experimental data of syngas solubility in n-decane, DMF, 1-dodecene, n-dodecanal as well as in mixtures of n-decane and DMF and in a mixture of all relevant liquids (1-dodecene, n

Acknowledgments

This work is part of the Sonderforschungsbereich/Transregio 63 “Integrated Chemical Processes in Liquid Multiphase Systems”. The authors thank the Deutsche Forschungsgemeinschaft (DFG) for financial support, New Ways of Analytics GmbH (Lörrach) for helpful advice on the view cell and Maria Schmidt and Rabea Kleinebrahm for performing some of the measurements.

References (18)

There are more references available in the full text version of this article.

Cited by (26)

  • Surrogate modeling of phase equilibrium calculations using adaptive sampling

    2019, Computers and Chemical Engineering
    Citation Excerpt :

    The accurate description of the solubility of the supercritical reactants hydrogen and carbon monoxide in the reaction mixture as well as the temperature-dependent LLE are crucial for predicting the performance of the process. During the investigation of this process, thermodynamic models accurately describing the behavior of all phase equilibria have been developed using PC-SAFT (Schäfer et al., 2012; 2014; Vogelpohl et al., 2013; 2014). The goal of our work is to perform process optimization based upon these models.

  • High-pressure solubility of light gases in heavy n-alkanes from a predictive equation of state: Incorporating Henry's law constant into binary interaction parameter

    2014, Fluid Phase Equilibria
    Citation Excerpt :

    Solubility of light gases in heavy hydrocarbons is important in coal liquefaction [1,2], petroleum processing [3,4], supercritical separation [5,6] and enhanced oil recovery [7–12].

View all citing articles on Scopus
View full text