Certain ant/oxidants added to polyethylene are known to cause increased dielectric loss in the audio-frequency range at liquid helium temperature. Although the peak loss factor occurs at several kilohertz, the increase at power transmission frequency is sufficient to affect the performance of an ac superconducting cable insulated with polyethylene tape. Contrary to recent reports, we now have reason to believe that the effect is an exclusive feature of antioxidant molecules which are 3, 5-ditert-butyl-4-hydroxy-phenyl compounds. It is attributable to tunnelling of the hydroxy hydrogen atom. Other substituted phenols and non-phenol ant/oxidants can all effectively protect the polyethylene without increasing the dielectric loss, 4.2 K dielectric loss measurement may be used as a rapid, sensitive and specific test for the presence of 3, 5-ditert-bu tyl-4-hydroxy-phenyl compounds in polyethylene,
Low temperature dielectric loss due to antioxidants in polyethylene J. le G. Gilchrist
In 1975, Thomas and King 1 reported that the dielectric loss factor of polyethylene at 4.2 K often exhibits a broadened relaxation peak at several kilohertz. This appeared to be associated with the presence of certain antioxidant additives. Ethyl 330 (see Fig. ) was cited as an example of an antioxidant having this property and BHT (3, 5-ditert-butyl-4hydroxy-toluene) as one that did not apparently cause increased dielectric loss. Yano et al2 confirmed these results and studied the temperature variations between 1.5 K and 4.2 K of the loss peak and its frequency. The peak loss factors reported by Thomas and King were around 1 -2 x 10 -4 but Yano et al apparently found loss peaks of 2 -3 x 10 -3. More recently we examined a series of low density polyethylene samples with various antioxidants, a'4 We classified the antioxidants into those containing the 3, 5-ditert-butyl-4hydroxy-phenyl group (alternatively the '2, 6-ditert-butylphenols'), other phenols and non-phenol antioxidants. How-~ ever, we failed to find a connection between the molecular structure of the antioxidant and the occurrence of the loss peak, other than a possible correlation with the molecular weight, and were unable to draw a firm conclusion.