I don't think that the hyperbranching as such matters a lot. The dielectric constant will be determined by the local mobilities of the polar parts of the molecule. If polar groups can easy fiddle around, the dielectric constant will be higher. Also the amount of aromatic groups will be an issue, because they can contribute to the induced dipole moment. Can you say something about the Tg's of both the linear and hyperbranched compounds? And are the chemical compositions comparable: the amount of aromatic groups, the number of polar groups, etc? And how big is the difference dielectric constant?

branching will have a considerable impact on the dielectric value of the polymer. So, too, will the moieties added which have produced the branching. I can speak from first-hand experience on this very point.

While I disagree with Dr. Huinink's first point, I wholly agreed with his second point. Further, do you have any substituents on the aryl moieties?

More pathways for q = lower dielectric value.

Thanks Stephen for your remark, because that sharpens it more. When I wrote "I don't think that the hyperbranching as such matters a lot", my intention was to shift the discussion from the emperical level (linear versus hyperbranched) to the molecular background of the dielectric constant. My point is that it is to simple to point branching as a cause if many other aspects on molecular scale change. I fully agree with that hyperbranching leads to a different dielectric constant.

@ H. Huinink, Sir, I could not find any Tg in DSC as it is low molecular weight resin. But the chemical compositions are comparable, linear epoxy is diglycidal ether of bisphenol-A (Dielectric constant 4.0-4.1) and the hyperbranched epoxy (Dielectric constant 1.9-2.1) is the polycondenced product of diglycidyl ether of bisphenol-A with a polyol (wt ratio 10:1). Both the linear and hyperbranched epoxy are cured with commercial vegetable oil based poly(amido-amine).

And the tests were 5-10 times repeated, but got same results

Polarizable fragments in hyperbranched molecules are ordered random (in equal proportion in all directions) by chemical bonds, so should have lower dielectric constant than linear polymer - with some provisos if thin film was used for measurement (surface ordering).

http://books.google.ru/books?id=800oLHINCIMC&lpg=PA64&ots=wTAaC9BorY&dq=bulk%20and%20thickness%20polymer%20film%20dielectric%20constant%20difference&hl=ru&pg=PA65#v=onepage&q=bulk%20and%20thickness%20polymer%20film%20dielectric%20constant%20difference&f=false

We have studied the dynamics of hyperbranched polymers and their linear analogue using dielectric spectrscopy and the results showed the dielctric constant is affected by branching, I hope this study can help you in your work.

A. Kyritsis, K. Raftopoulos, M. Abdel Rehim, Sh. Said Shabaan, A. Ghoneim, G. Turky, "Structure and molecular dynamics of hyperbranched polymeric systems with urethane and urea linkages", Polymer, 2009, 50, 4039

Thanks to all......and @ Andrei Rogoza, Sir, in our institute this book is not accessible, so please if you kindly can attach the the book.