Abstract:
Correlation study between ‘commencement of Nematic Transition Temperature (NTT)’ and their
‘structural property data set’ is carry out. Stepwise backward regression analysis method is applied to
find out good correlation between ‘transition temperature data set’ and ‘physical descriptors. Physical
descriptors are selected on the basis of good r2
-value and P-values with their respective NTTs. Data set
of randomly selected 32 compounds is used as training set to obtain “Quantitative Structure and Property
Relationship (QSPR) model”. Validation of derived QSPR equation is carried out on trial series as well
as on test series. Average standard deviation of 5.93 for 32 compounds of trial set, and 7.09 for 11
compounds of test series is observed; between predicted NTTs and experimentally measured NTTs.
Thus, it holds well on the data set of trial series and test series with comparative good degree of accuracy.
Hence, derived QSPR equation can utilize to design new similar type mesomorphs having comparative
low NTTs, so that they can be use at desired temperature.
Description:
We are especially thankful to Dr. Annasaheb Suryawanshi (Dept. of Statistics, Saurashtra University, Rajkot,
India), Dr. K. D. Ladva (Chemistry Department, Shree M. & N. Virani Science College, Rajkot, Gujarat, India)
and Dr. Devinder Tullysir (Sr. Scientist, TRC, Bhat, Gujarat, India) for giving me help and valuable suggestions
regarding to property calculations and regression analysis.
