No abstract

A new developing field of application for pressure swing adsorption (PSA) processes is the capture of CO 2 to mitigate climate change, especially the separation of CO 2 and H 2 in a pre-combustion context. In this process scheme the conditions of the feed to the separation step, namely a pressure of 3.5 to 4.5 MPa and a CO 2 fraction of around 40% are favorable for an adsorption based separation process and make PSA a promising technology. Among the commercial adsorbent materials, activated carbon is most suitable for this application. To evaluate the potential, to benchmark new materials, and for process development a sound basis of the activated carbon thermodynamic data is required, namely equilibrium adsorption isotherms of the relevant pure components and mixtures, Henry's constants and isosteric heats.In this work pure adsorption equilibria of CO 2 , H 2 and N 2 on commercial activated carbon (AP3-60 from Chemviron, Germany) are measured using a Rubotherm Magnetic Suspension Balance (MSB) (Bochum, Germany) in a wide temperature and pressure range. The data is used to fit the temperature dependent parameters of Langmuir and Sips (Langmuir-Freundlich) isotherms and to determine the Henry's constants as well as isosteric heats. Based on this evaluation different methods to evaluate the data are compared and discussed. With the pure isotherm parameters of the Sips isotherm binary adsorption is predicted using an empirical binary Sips equation and ideal adsorbed solution theory (IAST). The results are compared to binary measurements in the same MSB applying a gravimetricchromatographic method. Notation a parameter for temperature dependent description of n ∞ i [mmol/g] A specific surface of the adsorbent (Gibbs Adsorption isotherm) [m 2 /kg] A parameter for temperature dependent description of k i [1/MPa] b parameter for temperature dependent description of n ∞ i [J/mol] B parameter for temperature dependent description of k i [J/mol] c exponent in Sips isotherm [-] C parameter for temperature dependent description of H i [mmol/g/MPa] D parameter for temperature dependent description of H i [J/mol] E squared error from isotherm fitting [(mmol/g) 2 ] g weighting factor (final) [-] H Henry's constant [mmol/g/MPa] H heat of adsorption [kJ/mol] k isotherm equilibrium constant [1/MPa] m mass [g] M 1 weight at measuring point 1 [g] M 0 1 weight at measuring point 1 under vacuum [g] M molar mass [g/mol] n molar adsorption per unit mass of adsorbent [mmol/g] N number of experimental data points at one temperature (one component) [-] 50 Adsorption (2012) 18:49-65 p pressure [MPa] r weighting factor [-] R ideal gas constant [J/mol/K] S selectivity [-] t weighting factor [-] T Temperature [K] V Volume [cm 3 ] V 0 Volume of lifted metal parts and adsorbent [cm 3 ] V void Void volume of the adsorption system [cm 3 ] V 2 second virial coefficient for isotherm description [g/mmol] V 3 third virial coefficient for isotherm description [g 2 /mmol 2 ] w mass fraction [-] y mole fraction in fluid [-] z mole fraction in adsorbed ph...

No abstract

Pressure swing adsorption (PSA) experiments are carried out in a 2-column laboratory setup using activated carbon. As feed an equimolar CO2/H2 mixture is used. Adsorption pressures of 10, 20, and 30 bar are applied, whereas the desorption pressure and the feed temperature are fixed at 1 bar and 25 °C, respectively. During the experiments the temperatures at five different locations inside the columns are measured and the composition of the product streams is analyzed by a mass spectrometer. A one-dimensional, nonisothermal, nonequilibrium model is used to reproduce the experiments. The model was validated previously using breakthrough experiments, and the modifications required to describe full PSA cycles are highlighted. It is shown that the temperatures measured inside the columns provide an excellent possibility for comparison of experiments and simulations, whereas the measured concentration profiles are affected by the piping between column outlet and MS, which has to be considered carefully.

No abstract