Microsoft Word - 19Venkatesan.docx CHEMICAL ENGINEERING TRANSACTIONS VOL. 80, 2020 A publication of The Italian Association of Chemical Engineering Online at www.cetjournal.it Guest Editors: Eliseo Maria Ranzi, Rubens Maciel Filho Copyright © 2020, AIDIC Servizi S.r.l. ISBN 978-88-95608-78-5; ISSN 2283-9216 Prediction of Pure Component Properties of Alkenes and Dienes by Group Contributions Charles Otobrise Department of Chemistry, Delta State University, P.M.B. 1, Abraka, Nigeria otobrisec@delsu.edu.ng The pure component properties of many alkenes and dienes required for the design of process plants are not common in the open literature. Estimation methods are veritable tools for bridging this information gap. In this study, three Group Contribution Methods were employed to predict critical properties of twenty three alkenes and ten dienes. The capabilities of the three methods were compared and contrasted within obtainable experimental data. Generally, there was a good agreement between predicted and experimental properties. 1. Introduction Alkenes have a wide range of applications in the industry. They are used as precursors in the syntheses of detergents, plastics, alcohols, lacquers, and fuels. In particular they serve as feedstock for the petrochemical industry since they can take part in numerous reactions, especially polymerization and alkylation reactions. Pure component and thermodynamic property data of these compounds are crucial for the fabrication of process plants (Cunico et al., 2013). Quite a lot of industrially important compounds lack experimental data (Kolská et al., 2012), hence, different estimation methods have been developed over the years to provide data that cannot be sourced readily in the open literature (Constantinou and Gani, 1994). Such models are validated by comparing available experimental data with predicted values (Patrasciou and Doicin, 2015). Group Contribution Methods (GCMs) have been found to be very suitable and easy to employ for the prediction of a large number of pure component properties (Monago and Otobrise, 2010), as well as thermodynamic and physico-chemical properties of thousands of compounds (Marrero and Gani, 2001). GCMs have been developed to predict properties of hazardous compounds (Mondejar et al., 2017) and compounds that pose serious environmental challenges (Moosavi et al., 2014). He et al., (2016) evolved a Group Contribution Method (GCM) for the prediction of certain thermodynamic characteristics of hydrocarbon compounds with bulky carbon-carbon chain. The group interaction within a molecule was obtained by using an exponential decay function of the distance between atom groups in series. Overall, the GCM provided satisfactory results for alkanes and alkenes. Many GCMs have a limitation of being capable to predict properties of pure compounds, this is because information on the features of the disparate relations between groups of atoms and molecules or between atom groups alone is not easily extracted from pure component data, that are fundamental in determining the contributions of functional groups (Papaioannou et al., 2014). Three pure component properties are extensively used, namely: critical temperature (Tc), critical pressure (Pc) and critical volume (Vc) (Verevkin et al., 2006; Otobrise et al., 2018). The objective of this study is to predict with acceptable accuracy the critical properties of some alkenes and dienes, by means of GCMs. Experimental critical property data are mostly only obtainable for molecules of sufficient thermal stability. Molecules that are thermally labile often require estimation or prediction techniques to determine their pure component properties. 2. Estimation Methods Three GCMs were tested for critical property prediction, namely; the methods of Lydersen (1955), Constantinou and Gani (1994) and Marrero and Gani (2001). DOI: 10.3303/CET2080021 Paper Received: 15 November 2019; Revised: 9 February 2020; Accepted: 22 April 2020 Please cite this article as: Otobrise C., 2020, Prediction of Pure Component Properties of Alkenes and Dienes by Group Contributions, Chemical Engineering Transactions, 80, 121-126 DOI:10.3303/CET2080021 121 2.1. The Method of Lydersen The method utilizes contributions from atoms, groups of atoms or functional groups to estimate Tc, Pc and Vc. Equations 1, 2 and 3 gives the relationship for the prediction of critical properties. ( ) 1 2 0.567c b T TT T −  = + ΣΔ − ΣΔ  (1) ( ) 234.0 −ΣΔ+= pc MP (2) vcV ΣΔ+= 40 (3) In equation (1) Tc is in units of Kelvin; in equation (2) Pc is in units of atmospheres, however the final predictions were converted to bar units in tandem with the unit of the same parameter for other GCMs in this study. In equation (3) Vc is in units of cubic centimeters per mole. The summation of contributions for various atoms or groups of atoms yields the Δ quantities. Lydersen’s critical property values for various atoms or groups can be easily sourced (Reid et.al., 1977). The required experimental input parameters for this GCM are the normal boiling point (Tb) and the molar mass (M). The average relative deviations (ARD) associated with these relations varies. 2.2 The Method of Contsantinou and Gani (C & G). This GCM finds application in the estimation of a number of thermodynamic properties. The prescriptions for Tc, Pc and Vc are as follows: ( )  += i j jjiicoc DMWCNtT /exp (4) ( ) j j j j iiccc DMWCNPPP  +=−− − 25.01 (5) c coV -V i i j j i j N C W M D= +  (6) The parameters and summation quantities represented by the various symbols in equations (4) – (6) and contributions for various atoms or groups have been described elsewhere (Constantinou and Gani, 1994; Otobrise et al., 2018). 2.3 The Method of Marrero and Gani (M & G). This GCM estimates the properties of pure organic compounds in three steps. The prescriptions for Tc, Pc and Vc are as follows: ( )exp /c co i i j j k k i j k T t N C W M D Z O E= + +   (7) ( ) 0.51 2c c c i i j j k k j j k P P P N C W M D Z O E − − − = + +   (8) c coV -V i i j j k k i j k N C W M D Z O E= + +   (9) The summation quantities and other parameters, represented by the various symbols in equations (7) – (9) and contributions for various atoms or groups have been described elsewhere (Marrero and Gani, 2001; Otobrise et al., 2018). 122 3. Results and discussions In table 1 the predicted values of Tc, Pc and Vc for the alkenes and dienes by the GCMs are presented. The method of Constantinou and Gani under-predicted Vc unlike the other GCMs. The negative value of the adjustable parameter Vco in equation (9) makes this GCM particularly unsuitable for the prediction of critical volumes of alkenes and dienes. The prediction of the properties of isomers improved with GCMs that had second and third order contributions, especially the method of Marrero and Gani. Distinguishing the properties of isomers was particularly difficult with the method of Lydersen. In tables 2 and 3 the deviations of predicted properties from available experimental data of some alkenes and dienes are presented. Table1: Critical Properties of Alkenes and Dienes Predicted by the GCMs Alkenes Nc Predicted Tc/K Predicted Pc/bar Predicted Vc/cm 3 mol -1 Lydersen C & G M & G Lydersen C & G M & G Lydersen C & G M & G Ethene 2 280.65 292.04 271.08 48.56 65.99 61.87 140.00 112.13 106.38 1-Propene 3 363.15 344.32 371.24 43.32 49.48 49.14 195.00 187.17 162.66 1-butene 4 417.66 420.37 426.07 38.26 43.18 42.43 250.00 242.93 244.01 (E)-2-Butene 4 433.47 429.74 435.49 39.35 42.03 41.44 246.00 241.14 243.08 (Z)-2-Butene 4 439.64 430.03 436.42 39.35 71.81 68.37 246.00 69.71 243.58 2-methyl-1- propene 4 432.99 414.58 424.61 57.38 39.83 42.19 241.00 237.56 236.05 2-methyl-1-butene 5 475.02 469.48 469.15 33.98 35.28 36.98 296.00 293.32 292.33 2-methyl-2-butene 5 474.45 477.91 475.53 34.79 35.43 35.84 291.00 294.04 296.89 3-methyl-1-butene 5 451.99 456.08 460.34 34.79 38.00 37.60 301.00 293.74 312.36 1-Pentene 5 461.36 473.77 470.36 33.98 38.04 37.18 305.00 298.69 300.29 (E)-2-Pentene 5 476.96 480.79 478.17 34.79 37.09 41.44 301.00 296.9 299.36 (Z)-2-Pentene 5 477.87 481.01 478.94 34.79 71.81 56.98 301.00 125.47 299.86 2,3-Dimethyl-2- butene 6 540.82 524.41 524.00 31.09 32.01 32.01 337.00 371.99 372.01 3,3-Dimethyl-1- butene 6 479.78 489.67 495.19 31.09 34.64 33.60 346.00 336.91 340.44 1-Hexene 6 506.30 514.96 507.52 30.46 33.79 32.98 360.00 354.45 356.57 1-Heptene 7 490.92 548.49 539.52 31.49 30.25 29.58 415.00 410.21 412.85 1-Octene 8 563.32 576.78 567.63 25.13 27.25 26.79 470.00 465.97 469.13 2,4,4-Trimethyl-1- Pentene 8 559.02 556.50 557.35 25.53 26.14 27.10 447.00 443.06 445.04 2,4,4-Trimethyl-2- Pentene 8 561.78 561.76 561.73 32.65 26.23 26.45 442.00 443.78 389.25 1-decene 10 611.25 622.78 615.30 21.33 22.50 22.51 580.00 577.49 581.69 1-undecene 11 631.88 642.04 635.89 19.82 20.60 20.85 635.00 633.25 637.97 1-dodecene 12 650.49 659.44 654.80 18.51 18.95 19.43 690.00 689.01 694.25 1-hexadecene 16 711.84 715.98 717.95 14.62 14.09 15.40 910.00 912.05 919.37 Dienes 1,2-Butadiene 4 456.85 444.10 455.95 38.70 45.59 44.15 231.00 219.00 219.87 1,3-Butadiene 4 426.25 417.59 431.37 39.83 46.27 45.40 236.00 228.61 230.81 2-methyl-1,3- Butadiene 5 482.09 445.61 473.56 35.21 37.52 39.30 281.00 279.00 279.13 3-methyl-1,2- Butadiene 5 535.28 468.55 506.38 34.37 35.91 36.74 277.00 294.04 291.01 1,2-Pentadiene 5 497.71 491.73 495.30 34.37 40.02 38.53 286.00 274.76 276.15 (E)-1,3-Pentadiene 5 486.93 478.80 482.40 35.21 39.52 71.11 291.00 282.58 286.16 (Z)-1,3-Pentadiene 5 490.09 479.02 481.92 35.21 66.18 61.95 291.00 281.60 286.66 1,4-Pentadiene 5 462.10 471.70 451.60 35.21 40.57 39.51 291.00 284.37 287.09 2,3-Pentadiene 5 508.46 506.08 438.40 35.21 44.15 39.87 282.00 292.27 284.82 1,5-hexadiene 6 512.55 513.31 511.25 31.44 35.89 34.86 346.00 340.13 343.37 123 The deviations of estimated properties were calculated using the relationship in equation (10). The deviations in the predicted properties from experimental data are clearly related to the number of carbon atoms in each molecule. ( ) ( ) ( ) - Pr 100data data data Experimental edicted D Experimental = × (10) Table2: Deviations of Predicted Tc from Available Experimental Tc. Alkenes Nc Predicted Tc/K Experimental Tc/K Deviation (%) Lydersen C & G M & G Lydersen C & G M & G Ethene 2 280.65 292.04 271.08 282.30a 0.58 -3.45 3.97 1-Propene 3 363.15 344.32 371.24 365.60 a 0.67 5.82 -1.54 1-butene 4 417.66 420.37 426.07 419.60 a 0.46 -0.19 -1.54 (E)-2-Butene 4 433.47 429.74 435.49 428.60 a -1.14 -0.27 -1.61 1-Pentene 5 461.37 473.77 470.36 464.80 b 0.74 -1.93 -1.20 1-Hexene 6 506.30 514.96 507.52 504.00 b -0.46 -2.18 -0.70 1-Heptene 7 490.92 548.49 539.52 537.40 b 8.65 -2.07 -0.40 1-Octene 8 563.32 576.78 567.63 567.00 b 0.65 -1.73 -0.11 1-decene 10 611.26 622.78 615.30 616.00 b 0.77 -1.10 0.11 1-dodecene 12 650.49 659.44 654.80 658.00 b 1.14 -0.22 0.49 1-hexadecene 16 711.84 715.99 717.95 718.00 b 0.86 0.28 0.01 Source: a(Wick et al., 2000) b(Brauner et al., 2008). Table3: Deviations of Predicted Vc from Available Experimental Vc. Alkenes Nc Predicted Vc/cm 3mol-1 Experimental Vc/cm 3mol-1 Deviation (%) Lydersen C & G M & G Lydersen C & G M & G 1-butene 4 250.00 242.93 244.01 239.00* -4.60 -1.64 -2.10 2-methyl-1- propene 4 241.00 237.56 236.05 238.00* -1.26 0.18 0.82 2-methyl-1- butene 5 296.00 293.32 292.33 292.00* -1.37 -0.45 -0.11 3-methyl-1- butene 5 301.00 293.74 293.58 302.10* 0.36 2.77 2.82 1-Pentene 5 305.00 298.69 300.29 296.00* -3.04 -0.91 -1.45 1-Hexene 6 360.00 354.45 356.57 350.00* -2.86 -1.27 -1.88 1-Heptene 7 415.00 410.21 412.85 413.00* -0.48 0.68 0.04 1-Octene 8 470.00 465.97 469.13 472.00* 0.42 1.28 0.61 1-decene 10 580.00 577.49 581.69 585.00* 0.85 1.28 0.57 1-undecene 12 635.00 633.25 637.97 642.00* 1.09 1.36 0.63 1-dodecene 12 690.00 689.01 694.25 700.00* 1.43 1.57 0.82 Dienes 1,2-Butadiene 4 231.00 219.00 219.87 219.00* -5.48 0.00 -0.40 1,3-Butadiene 4 236.00 228.61 230.81 220.80* -6.88 -3.54 -4.53 3-methyl-1,2- Butadiene 5 277.00 294.04 291.01 291.00* 4.81 -1.04 0.00 1,2-Pentadiene 5 286.00 274.76 276.15 276.00* -3.62 0.45 -0.05 1,4-Pentadiene 5 291.00 284.37 287.09 276.00* -5.43 -3.03 -4.02 2,3-Pentadiene 5 282.00 292.27 284.82 295.00* 4.41 0.93 3.45 1,5-hexadiene 6 346.00 340.13 343.37 328.00* -5.49 -3.70 -4.69 Source: *(Zuas and Styarini, 2009). 124 Generally, the GCMs of C&G and M&G had a tendency to under predict Tc and Vc. Also their predictive capability improved with increase in the number of carbon atoms of each compound. The GCM of Lydersen could only differentiate between isomers in the prediction of Tc due to the variation in their boiling temperature (Tb). This however, was not the case in the prediction of Pc and Vc where the experimental boiling temperature (Tb) was not an input parameter. In such cases the same critical values of Pc and Vc were obtained for isomers. In order to validate the predicted values for Tc, experimental data available from other studies were used to appraise the average relative deviation (ARD) according to the relationship: % Deviation ARD N =  (11) The ARD for predicted Tc by the GCMs is 1.17 %, -0.64 % and -0.23 % for the methods of Lydersen, C &G and M&G respectively. The ARD for Predicted Vc is -1.51 %, -0.28 % and -0.53 % respectively for the three GCMs. In terms of ARD, the GCMs of C & G and M & G are comparable and both elicited good predictions of Tc and Vc. For the other critical property considered in this work, i.e. Pc, experimental data could not be sourced in the open literature. Analysis among the selected GCMs can only be carried out in such a case in relation to the observed trends of their predicted values. Figures 1 and 2 shows the estimated values for the critical pressures for alkenes and dienes as a function of their carbon chain length. Figure 1: Critical Pressures of Alkenes Estimated by different GCMs. Figure 2: Critical Pressures of Dienes Estimated by different GCMs. It can be observed from figures 1 and 2 that the three methods predicted similar Pcs for alkenes and dienes. As the carbon chain increased values predicted by the methods of C & G and M & G become closer. The inability of the method of Lydersen to distinguish between isomers, gave rise to noticeable variations in the curve. 4. Conclusions A data bank was generated using three GCMs to estimate critical properties such as Tc, Pc and Vc for important alkenes and dienes in the chemical industry. Experimental data that are available for such 0 20 40 60 80 0 5 10 15 20 P c /b a r Carbon Chain Lydersen C & G M & G 0 20 40 60 80 0 1 2 3 4 5 6 7 P c /b a r Carbon Chain Lydersen C & G M & G 125 properties are very few. In the course of this work only Tc for a few alkenes were found in the open literature. The method of M & G had comparable predictive advantage for the critical properties considered over the methods of Lydersen and C & G. In the largest part, good agreement between experimental data and predicted values was attained. References Brauner N., Cholakov G.St., Kahrs O., Stateva R.P., Schacham M., 2008, Linear QSPRs for Predicting Pure Compounds Properties in Homologus series, AIChE Journal, 54(4), 978-990. Constantinou L., Gani R., 1994, New Group Contribution Method for Estimating Properties of Pure Compounds, AIChE Journal, 40(10), 1697 – 1710. Cunico L.P., Ceriani R., Guirardello R., 2013, Estimation of Physical Properties of Vegetable Oils and Biodiesel, Using Group Contribution Methods, Chemical Engineering Transactions, 32, 535-540. DOI: 10.3303/CET1332090 He T., Li S., Chi Y., Zhang H., Wang Z., Yang B., He X., You X., 2016, An Adaptive Distance-based Group Contribution Method for Thermodynamic Property Prediction, Phys. Chem. Chem. Phys., 18, 23822-23830 Kolská Z., Zábranský M., Randová A., 2012, Group Contribution Methods for Estimation of Selected Physico- Chemical Properties of Organic Compounds, Thermodynamics–Fundamentals and Its Application in Science pp 135 – 162 Marrero J., Gani R., 2001, Group-Contribution Based Estimation of Pure Component Properties, Fluid Phase Equil., 183-184, 183-208 Monago K.O., Otobrise C., 2010, Estimation of Pure Component Properties of Fatty Acids and Esters, J. Chem. Soc. Nigeria, 35(2), 142-148. Mondejar M.E., Cignitti S., Abildskov J., Woodley J.M., Haglind F., 2017, Prediction of properties of New Halogenated Olefins using two Group Contribution Approaches. Fluid Phase Equilibria, 433, 79-96. Moosavi M., Sedghamiz E., Abareshi M., 2014, Liquid Density Prediction of Five Different Classes of Refrigerant Systems (HCFCs, HFCs, HFEs, PFAs and PFAAs) using the Artificial Neural Network-group Contribution Method, International Journal of Refrigeration 48, 188-200. Otobrise C., Monago K.O., Milicent U.I., 2018, Group Contribution Method for the Estimation of Critical Properties of Some Linear Aliphatic Dimethyl Esters of Dicarboxylic acids. FUW Trends in Science & Technology Journal, 3(2), 658 – 663. Papaioannou V., Lafitte T., Avendaño C., Adjiman C.S., Jackson G., Müller E.A., Galindo A., 2014, Group Contribution Methodology Based on the Statistical Associating Fluid Theory for Heteronuclear Molecules formed from Mie Segments, Journal of Chemical Physics, 140(5), 054107. Patrascioiu C., Doicin B., 2015, Property estimation of commercial ecological gasoline, Chemical Engineering Transactions, 43, 247-252 DOI: 10.3303/CET1543042 Reid R.C., Prausnitz J.M., Sherwood T.K., 1977, The Properties of Gases and Liquids. McGraw-Hill Book Company, New York, pp. 12-15. Verevkin S.P., Svetlana A.K., Vladimir N.E., Nikitin E.D., Popov A.P., Krasnykh E.L., 2006, Vapor Pressures, Enthalpies of vaporization, and Critical Parameters of a Series of Linear Aliphatic Dimethyl Esters of Dicarboxylic Acids. J. Chem. Eng. Data, 51, 1896 – 1905. Wick D.C., Martin M.G., Siepmann J.I., 2000,Transferable Potentials for phase Equilibra. 4-United-Atom Description of Linear and Branched Alkenes and Alkylbenzenes. The journal of Physical Chemistry B. 104(33), 8008 – 8016. Zuas O., Styarini D., 2009, A Quantitative Structure Property relationship (QSPR) Evaluation of Critical Volume of Unsaturated Hydrocarbon, Alkenes and Alkynes using Simple Connectivity Indices. Reaktor, 12(4), 260-267. 126