<<
>>

THE BASIC MAINTENANCE AND RESULTS OF WORK

In introduction tendencies in development of modelling of properties and behaviour of the molecules, caused by their structure, prospect of this direction are described.

In chapter 1 application of the theory of counts for ordering and transfer of isomers of benzol is discussed.

The theory of counts is widely applied in chemistry to the decision of theoretical and applied problems. The conclusion of isomers of replacement of benzol on the basis of theorem pojja is manufactured. Formulas of a cyclic index of group of substitutions and manufacturing function for definition of number of isomers of replacement of benzol are positioned. The review of the fundamental concepts connected with a theme of dissertational work is manufactured. Modern statistical methods of construction of QSPR-models, a separation process of objects of research on training and test groups are described, philosophy mnozhestvenno linearly-regressionnogo analysis are is short mentioned. Classification and the basic types of molecular descriptors on an example of the topological indexes used in QSPR-modelling is stated. The new topological index of centre distance with use of a matrix of distances is developed. This molecular descriptor is used as explanatory variable of size at creation of experimental group for modelling "structure-property" of investigated objects. The calculation formula includes summation of all elements of a matrix of distances and division of result into quantity of atoms of the carbon which is a part of connection.

Where N - quantity of atoms of carbon in connection, Dij - elements of a matrix of distances molecular column G, representing the least short distance between tops iи jграфа G.

In the second chapter the potential probability of construction qualitative QSPR - models for calculation of boiling point and enthalpy of formation of alkanes in a gas phase with use of a new index is investigated.

For the purpose of studying of the yielded problem the database (DB) on standard temperatures of boiling (Tkip) alkanes (72 connections) and on enthalpy of formation (АґНо (g, 29)) alkanes in a gas phase (47 connections) has been created. For statistical calculations software package Fathom Dynamic Data Software, Microsoft Excel and ActivStat is used. In work help data of physical and chemical characteristics of structures are used.

QSPR-models for definition Tkip and AfHo^gK) have been constructed and divided on two groups - control and experimental. The control group has included 8 models which have involved index Randicha (χ), Wiener's index (W) and index Balabana (J). An index of centre distance Dadtвыступил in parametre in experimental group which totals 7 equations. The statistical characteristics used for definition of models with higher prognosticheskoj by possibility, include factor of determination R2, correlation factor r and a root-mean-square standard deviation s. The Most qualitative QSPR - models are included in table 1. The QSPR-model for forecasting Tkip with use of a topological index of centre distance offers high accuracy.

About it speaks value of a standard deviation. Wiener's traditionally used index in QSPR-model for calculation AfHo^gK) has led to the best forecast. Fluctuation of sizes of a standard deviation in control group fluctuates from 9,8 kdzh/MOLE to 19,7 kdzh/MOLE and
Factor of correlation from 0,74 to 0,95. It is possible to conclude, that indexes of experimental group are entered in the yielded intervals, specifying on perspectivity of discriminating ability of an index of centre distance.

Table 1. Results of QSPR-modelling of alkanes.

QSPR-model R2 r s
Tkip = 55,337ln (W) - 121,17 0,9531 0,98 8,8 °s
Tkip =-0,360 (Dadti) + 18,11Dadti - 87,69 0, 946 0,97 7,84 °s
AgN ° (298к) (eksp) =-80W0,2382 0,9106 0,95 9,8 kdzh/MOLE
AgN ° (g, 298к) (eksp) =-68,08ln Dadti - 28,43 0,884 0,94 14,70 kdzh/MOLE

In drawing 1 correlation between experimental and settlement data of alkanes for the most perspective models is represented.

Fig. 1. A parity between experimental and predicted values Tkip of alkanes taking into account the logarithm of an index of Wiener, b) taking into account an index of centre distance in polynomial function; a parity between the experimental and predicted values. ∖1H ° (1,298 κ) alkanes) taking into account Wiener's index in the form of sedate function, d) taking into account the logarithm of an index of centre distance.

Check of statistical hypotheses is made, which have included the test of discriminating ability of an index of centre distance at modelling Tkip (the p-importance 0,4076) and ∆fH0 (l5298 κ) alkanes (the p-importance 0,7671).

Trerja the chapter is devoted studying of applicability of an index of centre distance for modelling of properties of cycloalkanes. Index Randicha (χ) and Wiener's modified index for odd rings and index Gutmana for even rings (RW/Sz) have acted as explanatory variables in the control group numbering 5 representatives. The topological index of centre distance is parametre in the experimental group consisting of 4 objects. The made database includes 24 connections.

The most qualitative QSPR-models of cycloalkanes are listed in table 2.

Table 2. Results of modelling Tkip of cycloalkanes.

QSPR-model R2 r S
Ткип=69,223ln (RW) - 190,6 0,9844 0,99 9,02°С
Tkip =-0,313 Dadti2 + 17,58 Dadti - 57,38 0,975 0,99 9,96°C

Statistical data of QSPR-models for calculation Tkip specify in more high quality of model from control group. Fluctuation of a standard deviation of control group between 9,02°С and 85,8°С and factor of correlation between 0,96 and 0,99 is observed. Modelling ∆fHO (g, 298K) cycloalkanes hardly gave in to modelling. The experimental group has higher statistical characteristics. Check of statistical hypotheses which have included the test of discriminating ability of an index of centre distance at modelling Tkip (the p-importance 0,5012) and ∆fHO (Γ, 298K) cycloalkanes (the p-importance 0,0226) is made.

QSPR-models for Tkip cycloalkanes are reflected in drawing 2:

Fig. 2. A parity between experimental and predicted values Tkip of cycloalkanes taking into account the logarithm of the modified index of Wiener, b) taking into account an index of centre distance in polynomial function

The fourth chapter describes check of efficiency of an index of centre distance at construction of QSPR-models for boiling point (Tpl), melting points (Tkip), a distribution number in system octanol - water (LogP) and enthalpies of formation in a gas phase (AHY ^^) polycyclic aromatic hydrocarbons. The control group has been constructed with application of index Randicha (χ), Wiener's index (W), sizes invarianta the column on the basis of matrix Laplasa (LEL (G)) and totals 31 object. The experimental group used an index of centre distance Dadtiи has made 5 records. For research of properties PAU physical and chemical sizes were are taken from a database of Milan Exploratory Group on Hemometrike and QSAR. The created database has included 82 connections. The most perspective models PAU are included in table 3.

Table 3. Results of modelling of properties PAU.

QSPR-model R2 r s
Тпл=0,2373 W + 31,523 0,5925 0,77 8,96°С
Tkip = 17,828LEL-17,391 0,9458 0,97 11,58°С
LogP = 1,461 ln (W) - 3,3881 0,9307 0,96 0,25
LogP = 2,466ln (DADTI) - 4,106 0,914 0,96 0,287
∆fH ° (r, 298K) =-0.8022 χ 2 + 19.535 χ - 43.386 0,864 0,93 4,74 kcal/MOLE
∆fH ° (r, 298K) (eksp) = 6,544 (Dadti) 0,568. 0,86 0,94 4,99 kcal/MOLE

QSPR - models from control group surpass the analogues from experimental group. However the range of fluctuations of standard deviation Tpl in control group reaches to 55,06°С. Standard deviation Tkip of models from experimental group rather highly. But such extreme value is not alien to objects from control group. For example, use of an index of Wiener has led to a standard deviation 20,84°С. A standard deviation of the most promising model from control group by calculation LogPнемного more low, than at its analogue from experimental group. But the control group has scatter in these indexes from 0,25 to 0,72. More high quality QSPR - models for calculation AfHo ^^ is inherent in control group. But all range of scatter of a standard deviation in control group rose to 7,82 kcal/MOLE. The standard deviation of experimental QSPR-model is located at the bottom border of an interval, being the optimal direction.

Perspective models are represented graphically in drawing 3.

Fig. 3. A parity of experimental and predicted values Tkip PAU taking into account LEL (G), b) taking into account Dadti; a parity between experimental and predicted values LogPПАУ) with the account ln (W), d) with the account ln (DADTI).

Fig. 4. Molecule PAU.

The QSPR-model is developed for definition of enthalpy of formation of the aromatic hydrocarbons consisting of benzene rings, connected by one and more general side. This model allows to predict enthalpy of formation for the molecules formed of connected benzene chains, forming to three horizontal numbers. As control group the values of analogues received with application of widely known algorithms are used. Comparative evaluation test of this model has yielded positive result and is presented in table 9. The sample of hypothetical structure is represented in drawing 4.

Where Dadti - a topological index of centre distance, N1, N2и N3 - quantity of benzene rings in the first, second and third numbers of the count accordingly.

As the standard for model evaluation test comparison with widely known used methods of calculations is made.

Table 4. Statistical characteristics of settlement models: r correlation factor between the experimental and predicted sizes and a standard deviation for all PAU applicability of models. s (kcal/MOLE).

Model r s Model r s
Benson 0,95 3,59 Armitazh 0,90 4,84
Moiseyev 0,84 6,08 QSPR-model (2) 0,99 2,17

Comparing statistical data of received model (2) with other predicted methods, good discriminating ability of an index of centre distance is obvious. Estimating results of calculations of enthalpy of formation of polycyclic aromatic hydrocarbons, it is possible to conclude,
That the received QSPR-model are entered in the general tendency with the conventional algorithms. Some fluctuation of values in the smaller or big side is noticed, that also is characteristic for other methods. There is no tendency to regular overestimate or understating of defined sizes of physical and chemical characteristics.

Table 5. Results of calculations of enthalpy of formation PAU in a gas phase with use of QSPR-model (2) and them soizmerenie with other methods.

PAU The molecular

The formula

ABOUT

h about

⅛?

SCH N

<

Kcal/mole QSPR-model AgN ° (g, 2 (eksp)
Benson Moiseyev Armitage QSPR-model
Benzol СбН6 19,7 17,8 0,1 -3,1 0,1 -1,9
Naphthalene С10Н8 35,9 35,1 0,1 0,0 0,1 -0,8
Phenanthrene С14Н10 49,6 47,9 0,4 0,0 0,4 -1,7
Piren С16Н10 53,9 58,5 1,3 8,2 1,2 4,6
Anthracene С14Н10 55,2 53,6 -3,0 0,0 -3,0 -1,6
Krizen С18Н12 64,5 61,8 -0,5 -1,8 -0,5 -2,7
Trif¾nil¾n С18Н12 65,5 64,3 -3,7 -7,8 -3,7 -1,2
Benz [and] anthracene С18Н12 70,0 67,3 -4,5 -1,8 -4,5 2,7
Benz [e] phenanthrene С18Н12 69,6 70,1 -3,4 -3,5 -4,5 0,5
Perylene С20Н12 76,3 75,3 -9,4 -6,2 -9,5 -1,0

Check of statistical hypotheses which have included the test of discriminating ability of an index of centre distance at modelling Tpl (the p-importance 0,3813) and АґН ° (g, 298К) PAU (the p-importance 0,0835) is made.

<< | >>
A source: FEDINA JULIA ALEKSEEVNA. QUANTITATIVE MODELS In CORRELATIONS "STRUCTURE - PROPERTY" HYDROCARBONS And THEIR SUBSTITUTED. The dissertation AUTOABSTRACT on competition of a scientific degree of a Cand.Chem.Sci. Tver - 2017. 2017

More on topic THE BASIC MAINTENANCE AND RESULTS OF WORK:

  1. THE BASIC RESULTS OF WORK
  2. THE BASIC RESULTS OF WORK
  3. THE BASIC CONCLUSIONS AND RESULTS OF WORK
  4. the Basic results of work and conclusions
  5. THE BASIC RESULTS OF WORK AND CONCLUSIONS
  6. THE BASIC RESULTS OF WORK AND CONCLUSIONS
  7. THE BASIC MAINTENANCE OF WORK
  8. THE BASIC MAINTENANCE OF WORK
  9. the Basic maintenance of work
  10. THE BASIC MAINTENANCE OF WORK
  11. THE BASIC MAINTENANCE OF WORK
  12. THE BASIC MAINTENANCE OF WORK
  13. THE BASIC MAINTENANCE OF WORK
  14. II. THE BASIC MAINTENANCE OF WORK
  15. the Basic maintenance of work
  16. the BASIC MAINTENANCE of WORK
  17. THE BASIC MAINTENANCE OF WORK
  18. THE BASIC MAINTENANCE OF WORK
  19. THE BASIC MAINTENANCE OF WORK