IA data based, boiling point estimation for alcohol isomers using molecular weight and carbon chain.
0 likes44 views
This document analyzes which is a better predictor of boiling point for alcohol isomers - molecular weight or number of carbon chains. Regression analysis was performed using molecular weight and carbon chains as independent variables to predict boiling point (dependent variable) for alcohols with 5-12 carbons. The regression equations were then used to predict boiling points for structural isomers of butanol, pentanol, and hexanol and percent errors calculated. Molecular weight predictions had 14-20% errors while carbon chain predictions were slightly better with 14-15% errors.
IA data based, boiling point estimation for alcohol isomers using molecular weight and carbon chain.
- 1. Which model is a better predictor, using molecular weight or number of carbon chain 2 or more independent variable (predictor) Is boiling point associated with molecular weight and carbon chains. Molecular weight or number of carbon chains – independent variables (predictor) Boiling point of alcohol isomers – dependent variable (outcome) Using Regression and Anova for analysis Independent variable Dependent variable Is molecular weight or number carbon chains a good predictor Independent variable Dependent variable Data for b/p from CRC Handbook. Click here data IA secondary data based – Regression analysis for boiling point estimation for alcohol isomers, (butanol, pentanol, hexanol) B/point = x1 (molecular weight) + intercept B/point = x1 (number of carbons) + intercept or Research Question Use 5 -12 carbon chains for regression model Use regression to predict b/p for isomers of alcohol, butanol, pentanol, hexanol. Find the % error using expt values with predicted values. Using molecular weight, 2nd order as predictor for b/p. Using carbon chain 2nd order as predictor for b/p. MF Number carbon Molecular weight b/p CH3OH 1 32.04 64.7 C2H5OH 2 46.09 78 C3H7OH 3 60.09 97 C4H9OH 4 74.12 117.7 C5H11OH 5 88.15 138 C6H13OH 6 102.16 157 C7H15OH 7 116.88 175
- 2. Isomerism Molecules with same molecular formula but diff arrangement of atom Two types of Isomerism Positional Chain Isomer Functional Gp Isomer C – C – C – C – OH C4H10O1 Structural Isomerism • Same molecular formula • Diff structural formula • Diff arrangement of atom Diff hydrocarbon chain skeleton • Same molecular formula • Same structural formula • Diff spatial arrangement of atom Stereoisomerism Hydrocarbon Chain Isomer Diff functional gp position Diff functional gp C – C – C – OH ׀ CH3 C – C – C –C ׀ OH C – C – C – C ׀ OH C – C – C – C ׀ OH C – C – C – O – C Optical Isomer Geometric Isomer Isomer Physical property Chemical property Structural isomer - Hydrocarbon chain - Functional gp position - Functional gp Different Different Different Similar Similar Different Geometrical isomer Different Similar Optical isomer Similar Similar Structural formula – arrangement atoms in molecule (2/3D) H H ׀ ׀ H - C – C – H ׀ ׀ H H CH3CH3 ethane Display full SF Condensed SF Ball/stick model Spacefilling Click here chemical search. Data source for b/p Data PubChem. Click here Data PubChem. Click here Data Handbook. Click here data
- 3. Diff functional gp position C4H9OH Structural Isomerism • Same molecular formula • Diff structural formula • Diff arrangement of atom Diff hydrocarbon chain skeleton Diff functional gp position Diff functional gp CH3-CH2-CH2-CH3 ׀ OH Butan–1-ol CH3-CH-CH2-OH ׀ CH3 2-Methylpropan-1-ol CH3 ׀ CH3-C-OH ׀ CH3 CH3-CH2-CH-CH3 ׀ OH 2-Methylpropan-2-ol Butan-2-ol CH3-CH2-CH2-O-CH3 Methoxypropane CH3-CH2-O-CH2-CH3 Ethoxyethane 7 structural isomers CH3-CH-O-CH3 ׀ CH3 CH3-CH2-CH2=CH2 C4H8 CH3-CH=CH-CH3 CH3-C=CH2 ׀ CH3 CH2 – CH2 ׀ ׀ CH2 - CH2 C4H9Br CH3-CH2-CH2-CH2 ׀ Br CH3-CH2-CH-CH3 ׀ Br CH3 ׀ CH3-C-Br ׀ CH3 CH3- CH-CH2 ׀ ׀ CH3 Br C6H14 CH3-CH2-CH2-CH2-CH2-CH3 CH3-CH-CH2-CH2-CH3 ׀ CH3 CH3-CH2-CH-CH2-CH3 ׀ CH3 CH3 ׀ CH3-C-CH2-CH3 ׀ CH3 CH3- CH- CH-CH3 ׀ ׀ CH3 CH3 Diff hydrocarbon chain skeleton Diff functional gp position Diff functional gp Diff hydrocarbon chain skeleton 4 structural isomers 4 structural isomers 5 structural isomers Diff hydrocarbon chain skeleton
- 4. Homologous Series Class Functional Suffix Example Formula Alcohol Hydroxyl - ol methanol CnH2n+1OH • member differ by CH2 gp • same functional group • similar chemical properties • chemical formula CnH2n+1OH • end with ol Number carbon IUPAC name Structure formula b/p 1 Methanol CH3OH 64.7 2 Ethanol CH3CH2OH 78 3 Propanol CH3CH2CH2OH 97 4 Butanol CH3(CH2)2CH2OH 117.7 5 Pentanol CH3(CH2)3CH2OH 138 methanol ethanol propanol butanol H ׀ H - C – OH ׀ H H H ׀ ׀ H - C – C – OH ׀ ׀ H H H H H ׀ ׀ ׀ H - C – C – C – OH ׀ ׀ ׀ H H H H H H H ׀ ׀ ׀ ׀ H - C – C – C – C – OH ׀ ׀ ׀ ׀ H H H H Hydrocarbon skeleton Functional gp b/p increase ↑ Physical properties • Increase RMM / molecular size •RMM increase ↑ - Van Der Waals forces stronger ↑ ↓ boiling point increases ↑ (Increasing polarisability ↑) London dispersion forces/temporary dipole ↑ Number carbon Molecular weight b/p 1 32.04 64.7 2 46.09 78 3 60.09 97 4 74.12 117.7 5 88.15 138 6 102.16 157 7 116.88 175 8 130.23 195 9 144.26 214 10 158.28 230 11 172.31 243 12 186.34 260 14 214.39 289 15 228.41 299 17 256.5 308 19 284.5 345 Boiling point for diff alcohol boiling point increase with increase carbon atoms
- 5. MW 2nd order as predictor for b/p for alcohol CC 2nd order as predictor for b/p for alcohol Research Question Use 5- 12 carbon chains for regression model Use regression to predict b/p for structural isomers of alcohol (butanol, pentanol, hexanol) Using molecular weight (MW) 2nd order as predictor for b/p. Using carbon chain (CC) 2nd order as predictor for b/p. y = -0.0021x2 + 1.8371x - 8.0951 R² = 0.999 0 100 200 300 0 50 100 150 200 b/p molecular weight molecular weight vs b/p y = -0.4405x2 + 24.964x + 23.44 R² = 0.9992 0 100 200 300 0 2 4 6 8 10 12 14 b/p carbon chain carbon chain vs b/p Can both model be used to predict b/p for isomer of alcohol (Butanol, pentanol hexanol) Structural Isomers 4 carbons b/p predict MW (% error) predict CC (% error) 1- butanol 117.7 2- butanol 99 2- methyl propan-1-ol 108 2-methyl propan-2-ol 82 Min b/p 102 Max b/p 138 Mean b/p 102 Structural Isomers 5 carbons b/p predict MW (% error) predict CC (% error) 1- pentanol 138 2- pentanol 119 3 – pentanol 116 2-methylbutan-1-ol 129 3-methylbutan-1-ol 131 2-methylbutan-2-ol 102 3-methylbutan-2-ol 112 2,2-dimethylpropanol 113 Min b/p 102 Max b/p 138 Mean b/p 120 Structural Isomers 6 carbons b/p predict MW (% error) predict CC (% error) 1- hexanol 157 2- hexanol 140 3 – hexanol 135 2,3 –dimethylbutan-2-ol 118 3,3 –dimethylbutan-1-ol 136 2-methylpentan-1-ol 149 3-methylpentan-1-ol 153 4-methylpentan-1-ol 151 2-methylpentan-2-ol 121 3-methylpentan-2-ol 134 4-methylpentan-2-ol 131 2-methylpentan-3-ol 126 3-methylpentan-3-ol 122 Min b/p 118 Max b/p 157 Mean b/p 136
- 6. Predicted b/p for carbon 4 – MW of 74.12 2nd order fit, y = -0.0021x2 + 1.837x – 8.095 b/p= -0.0021(74.12)2 + 1.837(74.12) – 8.095 = 116 % error = (𝑬𝒙𝒑𝒕 𝒗𝒂𝒍𝒖𝒆 −𝑷𝒓𝒆𝒅𝒊𝒄𝒕𝒆𝒅 𝒗𝒂𝒍𝒖𝒆) 𝑬𝒙𝒑𝒕 𝒗𝒂𝒍𝒖𝒆 x 100% % error = (𝟏𝟎𝟐−𝟏𝟏𝟔) 𝟏𝟎𝟐 x 100% = 14% Research Question Use 5- 12 carbon chains for regression model Use regression to predict b/p for structural isomers of butanol (4 carbons) Using molecular weight (MW) 2nd order as predictor for b/p. Using carbon chain (CC) 2nd order as predictor for b/p. Structural Isomers 4 carbons b/p predict using MW (% error) predict using CC (% error) 1- butanol 117.7 2- butanol 99 2- methyl propan-1-ol 108 2-methyl propan-2-ol 82 Min b/p 102 Max b/p 138 Mean b/p 102 116 (14%) 116 (14%) Predicted b/p for carbon 4 2nd order fit, y = -0.44x2 + 24.96x + 23.44 b/p= -0.44(4)2 + 24.96(4) + 23.44 = 116 All isomers 4 carbon alcohol have same MW – 74.12
- 7. Predicted b/p for carbon 5 – MW of 88.15 2nd order fit, y = -0.0021x2 + 1.837x – 8.095 b/p= -0.0021(88.15)2 + 1.837(88.15) – 8.095 = 145 % error = (𝑬𝒙𝒑𝒕 𝒗𝒂𝒍𝒖𝒆 −𝑷𝒓𝒆𝒅𝒊𝒄𝒕𝒆𝒅 𝒗𝒂𝒍𝒖𝒆) 𝑬𝒙𝒑𝒕 𝒗𝒂𝒍𝒖𝒆 x 100% % error = (𝟏𝟐𝟎−𝟏𝟒𝟓) 𝟏𝟐𝟎 x 100% = 20% Research Question Use 5- 12 carbon chains for regression model Use regression to predict b/p for structural isomers of pentanol (5 carbons) Using molecular weight (MW) 2nd order as predictor for b/p. Using carbon chain (CC) 2nd order as predictor for b/p. Structural Isomers 5 carbons b/p predict using MW (% error) predict using CC (% error) 1- pentanol 138 2- pentanol 119 3 – pentanol 116 2-methylbutan-1-ol 129 3-methylbutan-1-ol 131 2-methylbutan-2-ol 102 3-methylbutan-2-ol 112 2,2-dimethylpropanol 113 Min b/p 102 Max b/p 138 Mean b/p 120 145 (20%) 137 (14%) Predicted b/p for carbon 5 2nd order fit, y = -0.44x2 + 24.96x + 23.44 b/p= -0.44(5)2 + 24.96(5) + 23.44 = 137 All isomers 5 carbon alcohol have same MW – 88.15
- 8. Predicted b/p for carbon 6 – MW of 102.16 2nd order fit, y = -0.0021x2 + 1.837x – 8.095 b/p= -0.0021(102.16)2 + 1.837(102.16) – 8.095 = 157 % error = (𝑬𝒙𝒑𝒕 𝒗𝒂𝒍𝒖𝒆 −𝑷𝒓𝒆𝒅𝒊𝒄𝒕𝒆𝒅 𝒗𝒂𝒍𝒖𝒆) 𝑬𝒙𝒑𝒕 𝒗𝒂𝒍𝒖𝒆 x 100% % error = (𝟏𝟑𝟔−𝟏𝟓𝟕) 𝟏𝟓𝟕 x 100% = 15% Research Question Use 5- 12 carbon chains for regression model Use regression to predict b/p for structural isomers of hexanol (6 carbons) Using molecular weight (MW) 2nd order as predictor for b/p. Using carbon chain (CC) 2nd order as predictor for b/p. Structural Isomers 6 carbons b/p predict using MW (% error) predict using CC (% error) 1- hexanol 157 2- hexanol 140 3 – hexanol 135 2,3 –dimethylbutan-2-ol 118 3,3 –dimethylbutan-1-ol 136 2-methylpentan-1-ol 149 3-methylpentan-1-ol 153 4-methylpentan-1-ol 151 2-methylpentan-2-ol 121 3-methylpentan-2-ol 134 4-methylpentan-2-ol 131 2-methylpentan-3-ol 126 3-methylpentan-3-ol 122 Min b/p 118 Max b/p 157 Mean b/p 136 157 (15%) 157 (15%) Predicted b/p for carbon 6 2nd order fit, y = -0.44x2 + 24.96x + 23.44 b/p= -0.44(6)2 + 24.96(6) + 23.44 = 157 All isomers 6 carbons alcohol have same MW – 102.16
- 9. carbon chain 2nd order model molecular weight 2nd order model % error 2nd order for both model are relatively the same. 2nd order fit – % error changes from 14% to 14% to 15% as carbon chain changes from 4 to 5 to 6. Both model produced about 14 -15% error. y = -0.4405x2 + 24.964x + 23.44 R² = 0.9992 0 50 100 150 200 250 300 0 2 4 6 8 10 12 14 b/p carbon chain carbon chain vs b/p y = -0.0021x2 + 1.8371x - 8.0951 R² = 0.999 0 50 100 150 200 250 300 0 50 100 150 200 b/p molecular weight molecular weight vs b/p Structural Isomers 4 carbons b/p predict using MW (% error) predict using CC (% error) Mean b/p 102 116 (14%) 116 (14%) Structural Isomers 5 carbons b/p predict using MW (% error) predict using CC (% error) Mean b/p 120 145 (20%) 137 (14%) Structural Isomers 6 carbons b/p predict using MW (% error) predict using CC (% error) Mean b/p 136 157 (15%) 157 (15%) Research Question Use 5- 12 carbon chains for regression model Use regression to predict b/p for structural isomers of alcohol (butanol, pentanol, hexanol) Using molecular weight (MW) 2nd order as predictor for b/p. Using carbon chain (CC) 2nd order as predictor for b/p.
- 10. carbon chain 2nd order model molecular weight 2nd order model % error 2nd order for both model are relatively the same. % error using CC appears to be smaller compared to MW model. % error small 2-4% for non aromatic cyclic OH. Both model produced about 11 -19% error for aromatic OH. y = -0.4405x2 + 24.964x + 23.44 R² = 0.9992 0 100 200 300 0 2 4 6 8 10 12 14 b/p carbon chain carbon chain vs b/p y = -0.0021x2 + 1.8371x - 8.0951 R² = 0.999 0 100 200 300 0 50 100 150 200 b/p molecular weight molecular weight vs b/p Molecules with OH Molecular weight Number carbon b/p predict using MW (% error) predict using CC (% error) phenol 94.11 6 181.7 146 (19%) 157 (13%) 3 methylphenol 108.14 7 203 166 (18%) 177 (13%) 2 phenylethanol 122.16 8 219 185 (16%) 195 (11%) Cyclohexanol 100.15 6 161 154 (4%) 157 (2%) Research Question Use 5- 12 carbon chains for regression model Use regression to predict b/p for aromatic molecules with OH gps. Using molecular weight (MW) 2nd order as predictor for b/p. Using carbon chain (CC) 2nd order as predictor for b/p. Phenol 3 methylphenol 2 phenylethanol cyclohexanol