Modelling the diffusion coefficient D for water as solvent

1 Modelling D as a function of MB. T, tjs, (CMR)A The model defined in Section 15.3.3 was fitted for water as solvent, under three sets of conditions. The first set was developed by ignoring viscosity rjB. Under this approach the model becomes  

The model was fitted using least squares regression which gave the following coefficient estimates and R2 value  

Although this model gives a high R2, ignoring rjB is contrary to the previous work described in the Introduction. 

The second model used observed viscosity rjB as an input. Since there was a reasonable amount of data for the water solvent we can include observed values of ijB in the model. Thus the model (with MB fixed) becomes  

The least squares coefficient estimates and R2 value for this model were  

---