Estimation of Vapour Pressures of Organic Liquids using Group Contributions and Group Interactions

Transcription

1 Estimation of Vapour Pressures of Organic Liquids using Group Contributions and Group Interactions B. Moller*, J. Rarey**, D. Ramjugernath* * University of Kwazulu-Natal, Durban 4041, South Africa ** Technische Chemie (F5), Universität Oldenburg, Oldenburg ICCT Warsaw

2 Overview Motivation Project Timeline Properties of the Vapor Pressure Curve Performance of Different Equations Model Description Software Tools Results Future Work Acknowledgement ICCT Warsaw

3 P (kpa) What are Important Properties? Usually need to know information about the vapour-liquid boundary, useful for example in: Distillation applications Environmental applications 1 E+05 1 E+03 T m T c 1 E+01 1 E-01 T t T b Water 1 E-03 1 E T (K) ICCT Warsaw

4 Project Timeline Rarey T b W. Cordes, J. Rarey, A New Method for the Estimation of the Normal Boiling Point of Non-Electrolyte Organic Compounds, Fluid Phase Equilibria, 201/2, (2002). ICCT Warsaw

5 Project Timeline Rarey T b New T b Y. Nannoolal, J. Rarey, D. Ramjugernath, W. Cordes, Estimation of Pure Component Properties, Part 1: Estimation of the Normal Boiling Point of Non-Electrolyte Organic Compounds via Group Contributions and Group Interactions, Fluid Phase Equilibria, 226, 45-63, ICCT Warsaw

6 Project Timeline Rarey T b New T b T c, P c, V c Nannoolal, Y., Rarey, J., Ramjugernath, D., Estimation of Pure Component Properties Part 2: Estimation of Critical Data by Group Contribution., Fluid Phase Equilib., 252 (2007) 1. ICCT Warsaw

7 Project Timeline Rarey T b New T b T c, P c, V c P s (T) Nannoolal, Y., Rarey, J., Ramjugernath, D., Estimation of Pure Component Properties Part 3: Estimation of the Vapour Pressure of Non-Electrolyte Organic Compounds via Group Contributions and Group Interactions., Fluid Phase Equilib. accepted for publication. ICCT Warsaw

8 Project Timeline Rarey Nannoolal T b New T b T c, P c, V c P s (T) μ(t) Nannoolal, Y., Rarey, J., Ramjugernath, D., Estimation of Pure Component Properties Part 4: Estimation of the Liquid Viscosity of Non-Electrolyte Organic Compounds via Group Contributions and Group Interactions., Fluid Phase Equilib. in preparation. ICCT Warsaw

9 Project Timeline Rarey Nannoolal Moller T b New T b T c, P c, V c P s (T) μ(t) New P s (T) Moller, B., Rarey, J., Ramjugernath, D., Estimation of the Vapour Pressure of Non-Electrolyte Organic Compounds via Group Contributions and Group Interactions., J. Mol. Liq. accepted for publication. ICCT Warsaw

10 Methods for Predicting Physical Properties Group contribution methods Fairly simple Good accuracy Need groups to make predictions Quantum chemistry (COSMO-RS, COSMO-SAC ) Great potential but not at the required level of accuracy Statistical thermodynamics (molecular simulations ) Can get good results when fitted to some reference data points ICCT Warsaw

11 Purpose of this Work A large number of good group contribution methods already developed Objective of this work was to rework the vapour pressure method Nannoolal based his vapour pressure method on the Antoine equation, the first step was therefore to review this decision ICCT Warsaw

12 Hvap (kj/mol) The Slope of the Vapor Pressure Curve Zvap Hvap/Zvap (kj/mol) S dlnp 1 1 R d T H Z vap vap T r Component: Benzene ICCT Warsaw T r

13 Approximate Modeling of the Slope Hvap/Zvap (kj/mol) a) 10 kpa T b 1000 kpa 3000 kpa b) Benzene H Z vap vap T r T r T T c a) Linear with respect to: 1/T b) Constant ICCT Warsaw

14 The Antoine Equation B logp s A T C 36 Hvap/Zvap (kj/mol) T b Benzene Deviation: 3.6% T r ICCT Warsaw

15 The Cox Equation s P Tb log A' 1 atm P T log A' log A E(1 T )( F T ) c r r 36 Hvap/Zvap (kj/mol) Benzene T r T T c 30 T b T r ICCT Warsaw

16 Kinetic Theory of Vaporization Hvap/Zvap (kj/mol) B lnp A C T DT ET T 2 0 ln A s s T b T b Benzene T r R E ln ( 0.5)ln ln ( ) ln Vw R E B 0 s 1 C 1.5 s D R E 0 R 2( s3)( s1) E E R E o s 2 - Characteristic energy - Number of loosely coupled harmonic oscillators Applicable to pressures lower than 200 kpa 0 Abrams, D. S., Massaldi, H. A., Prausnitz J. M., Vapor Pressures of Liquids as a Function of Temperature. Two-Parameter Equation Based on Kinetic Theory of Fluids., Ind. Eng. Chem. Fundam. 13(3) (1974) 259. Moelwyn-Hughes, E. A., The Chemical Statics and Kinetics of Solutions, Academic Press, London (1971). ICCT Warsaw

17 Model Derivation The A parameter in the Antoine equation is substituted by using the normal boiling point as a reference point: B B lnp lnp atm T C Tb C In order to further reduce the number of model parameters it is assumed that C is a function of the normal boiling point: ln P T Tb B' Patm T C( Tb ) For aliphatic alcohols and carboxylic acids the new form is: P T T b T ln B' D'ln Patm T C( Tb ) Tb ICCT Warsaw

18 Model Equations P T T b T ln B' D'ln Patm T C( Tb ) Tb (55 GI s) n n 1 B' A db GIi i1 j1 (158 Groups) j Tb CT ( b ) Aliphatic alcohols & carboxylic acids m 1 D' D vide n a i 1 (5 Groups) i db v db n v db db i i a j j k i j k Group interactions (H-bonding groups) Frequency Existence Size n a v - number of atoms except H - frequency of a group ICCT Warsaw

19 Group Contributions - db i as Function of the Number of Heavy Atoms dbi alkyne-groups dbi aliphatic alcohol groups number of atoms number of atoms dbi aliphatic carboxylic acid groups Short and longer chain alcohols and carboxylic acids require separate groups number of atoms ICCT Warsaw

20 P (kpa) P (kpa) Model Improvements Alcohol and carboxylic acid representation Smaller and larger molecules Low pressure data Current work Nannoolal /T (K -1 ) T (K) 1-Nonanol (data from the DDB) ICCT Warsaw

21 Absolute Error (K) Model Parameters Error Relationship Error for the prediction of the boiling point increases with number of model parameters! Gani, Constantinon Marrero, Pardillo EWOR Stein, Brown Cordes, Rarey Number of parameters EWOR = Ericksen, Wilding, Oscarson, Rowley ICCT Warsaw

22 Simplification of Group Definitions Replace specific groups with more widely applicable groups: Example: double bonded carbon group: Previously (Nannoolal et al.) New method C C HC CH C H C C H 2 C CH CH H 2 C C H 2 C CH 2 CH 2 6 groups to describe all combinations Only 3 groups required to describe all combinations These simplifications may not be applicable to the estimation of other properties! ICCT Warsaw

23 Introduction of New Structural Groups aromatic iso-cyanate hydrazine cyclic tertiary amines halo-silicon groups (F,Cl,Br,I) fused ring carbon (and other structural groups) silicon group expanded ICCT Warsaw

24 Software Tools for Group Analysis ICCT Warsaw

25 Vapour Pressure Data Dortmund Data Bank (DDB) VAP data for compounds VAP points Data needed to be validated Used VBA to streamline the process ICCT Warsaw

26 Data Validation ICCT Warsaw

27 Results for All Available Data Relative mean deviations (%) for the different vapour pressure ranges exp P P RMD(%) 100% exp P This work Nannoolal et al. (NC = 2332) (NC = 2207) Average (All Pressures) HP (>500 kpa) MP (10 kpa kpa) LP (10 Pa - 10 kpa) ELP (<10 Pa) NC number of compounds ICCT Warsaw

28 Detailed Results (Relative Mean Deviation (%)) Group NC ELP LP MP HP total All oxygen compounds Carboxylic acids Aromatic carboxylic acids Aliphatic carboxylic acids Alcohols Aromatic alcohols Aliphatic alcohols Ethers Esters Ketones Aldehydes Carbonate diesters Anhydrides Epoxides Carbonates Ureas NC number of compounds ELP extremely low pressure P < 10 Pa LP low pressure 10 Pa < P < 10 kpa MP medium pressure 10 kpa < P < 500 kpa HP high pressure P > 500 kpa total whole vapour pressure range ICCT Warsaw

29 Quality Analysis Can use the error tables to give an approximate quality of prediction: Compound to be predicted Give quality analysis based on training set errors Select most specific group which the compound falls into (e.g. n-alcohols) ICCT Warsaw

30 Number of Compounds RMD(%) Histogram for the Data in the Training Set % 7% % 31% % >60 RMD (%) Typically compounds with large amounts of low vapour pressure data for example squalane (C30H62) ICCT Warsaw

31 Results for a Test Set not Used in the Regression New Method Nannoolal et al. Compound Class NC RMD% NC RMD% Hydrocarbons Halogen compounds Oxygen compounds Nitrogen compounds Sulfur compounds All compounds Test set contains data with higher scatter or where the reliability could not be verified (only one author, ) ICCT Warsaw

32 P (kpa) Specific Test Set Examples 100 O OH 10 4-Methyl-4-hydroxy-2-pentanone /T (K -1 ) ICCT Warsaw

33 P (kpa) Specific Test Set Examples 100 OH 10 O alpha-hydroxy-isobutyric acid,methyl ester O /T (K -1 ) Danov S.M., Obmelyukhina T.N., Chubarov G.A., Balashov A.L., Dolgopolov A.A., J.Appl.Chem.USSR, 63(3), , 1990 Matin N.B., Khitrin S.V., Spasskaya R.I., Trachenko V.I., Zilberman E.N., Soluyanova T.F., Zh.Prikl.Khim.(Leningrad), 53(9), , 1980 ICCT Warsaw

34 P (kpa) Specific Test Set Examples 100 HO O F F 10 F Trifluoroacetic acid /T (K -1 ) ICCT Warsaw

35 Heat of Vaporization at K C( Tb) Tb Hvap RZ vap B ' D' T 2 CT ( b) 1 T This Work Kolská et al.* Compound Class NC RMD(%) NC RMD(%) Hydrocarbons Halogen Compounds Oxygen Compounds Nitrogen Compounds Phosphorous Compounds Sulphur Compounds Metals Other Compounds Silicon Compounds All Compounds *Kolská Z., Růžička V., Gani R., Ind. Eng. Chem. Res., 44, , 2005 ICCT Warsaw

36 Future Work Prediction of further properties Ground work has been done Overlap project so there is a transfer of knowledge Prediction of mixture properties ICCT Warsaw

37 Acknowledgements Supervisor: Prof. J. Rarey Project Leaders: Prof. D. Ramjugernath Prof. J. Gmehling Software: DDBST GmbH ICCT Warsaw