done) system with a headspace-gas chromatography (HSGC) apparatus at 333.15 K. The ionic liquids investigated were 1,3-dimethylimidazolium tetrafluoroborate [C2MIM]+[BF4]−, 1-butyl-3-methylimidazolium tetrafluoroborate [C4MIM]+[BF4]−, 1-methyl-3-octylimidazolium tetrafluoroborate [C8MIM]+[BF4]−, 1,3-dimethylimidazolium dicyanamide [C2MIM]+[N(CN)2]− and 1-octylquinolinium bis(trifluoromethylsulfonyl)amide [C8Chin]+[BTA]−. It was found that at low feeding concentration of 1-hexene and n-hexane, the separation ability of ionic liquids is in the order of [C2MIM]+[BF4]− > [C4MIM]+[BF4]− ≈[C2MIM]+[N(CN)2]− > [C8MIM]+[BF4]− > [C8Chin]+[BTA]−, which is consistent with the priori prediction of the COSMO-RS (conductor-like screening model for real solvents) model. But at high feeding concentration, the separation ability of ionic liquids is in the order of [C2MIM]+[BF4]− < [C4MIM]+[BF4]− ≈[C2MIM]+[N(CN)2]− < [C8MIM]+[BF4]−< [C8Chin]+[BTA]−. The liquid demixing effect should be taken into account. The activity coefficients of 1-hexene and n-hexane at infinite dilution calculated with the COSMO-RS model were correlated using the NRTL, Wilson and UNIQUAC model. In this work the predictive results from the COSMO-RS model and UNIFAC model for the 1-hexene/n-hexane and 1-hexene/n-hexane/NMP systems were compared. The UNIFAC model is one of the most important academic contributions by Prof. Jurgen Gmehling.

This work deals with the mathematic models and improvement of four kinds of gas-liquid and gas-liquid-solid reactors for producing cumene by catalytic reaction, i.e. fixed-bed reactor, fixed-bed catalytic distillation (FCD) column, suspension catalytic distillation (SCD) column and ionic-liquid catalytic distillation (ICD) column. The fixed-bed reactor was improved by changing the particle diameter of catalyst and no extra equipment was needed. The FCD and SCD columns were respectively improved by carrying out alkylation and transalkylation simultaneously in a column, which also may hold for the ICD column, in which, however, the COSMO-RS (conductor-like screening model for real solvents) model was used to take into account the influence of ionic liquid on VLE (vapor-liquid equilibrium).

This work presents the first research linking chemical engineering and sport science as far as we know. The COSMO-RS (conductor-like screening model for real solvents) model was used to make a priori prediction for the extraction of stimulants from aqueous solution by the double action of supercritical carbon dioxide (SC CO2) and ionic liquid. It was found that the suitable ionic liquids should have small molecular volume, unbranched group and no sterical shielding effect around anion charge center, and thus [C2MIM]+[OAc]− is the best among all the ionic liquids investigated. The calculated results from the COSMO-RS model were qualitatively consistent with those from experiments. On this basis, partition coefficients of amphetamine (C9N) and nikethamide (C10N) between aqueous phase and supercritical fluid (or MTBE) phase at different temperatures were calculated. It was shown that the separation efficiency of supercritical extraction with ionic liquid is generally higher than that of traditional liquid–liquid extraction. The modeling present can also be extended to the separation of trace amount of organic substances from aqueous solutions for other purposes.

Solubilities are reported for carbon dioxide (CO2) in polypropylene (PP) at temperatures from 313.2 to 483.7K and at pressures up to 25MPa. Buoyancy corrections required in the data reduction were estimated with the Sanchez and Lacombe equation of state (S-L EOS). Solubilities of CO2 in molten state PP could be correlated to within 5% for both the S-L EOS and the group-contribution lattice-fluid equation of state (GCLF EOS). The change in crystallinity of rubbery state PP with CO2 dissolution into the polymer was predicted with the GCLF EOS and the solubility data under the assumption that CO2 could dissolve only in the amorphous regions. The trend of the estimated crystallinity with CO2 dissolution into the polymer was consistent with results in the literature.

This work deals with the separation of 1-hexene and n-hexane as the representation of olefins and paraffins with ionic liquids, as well as N-methyl-2-pyrrolidone (NMP) screened by computer-aided molecular design (CAMD). On the basis of conformation analysis of solvents and ionic liquids, the conductor-like screening model for real solvents (COSMO-RS) was used to make a priori prediction for suitable ionic liquids. It was found that the suitable ionic liquids should have small molecular volume, unbranched group and sterical shielding effect around anion charge center. Headspace-gas chromatography (HSGC) experiments were done at 313.15 and 333.15 K. It was verified that the anion with sterical shielding effect around anion charge center is favorable for increasing the selectivity, and [C8Chin]+[BTA]− is the best among all the ionic liquids investigated. The separation mechanism of olefins and paraffins with ionic liquids can be explained by the theory of Prausnitz and Anderson’s solution thermodynamics. This work also can be extended to the separation of other hydrocarbons with ionic liquids since the separation mechanism between n-hexane/1-hexene and other hydrocarbons is consistent.