Comment on “Terahertz time-domain spectroscopy of films fabricated from SU-8”

For the original paper see ibid., vol. 35, no. 3, p. 243-44 (1999). The negative photoresist NANO^TM XP SU-8 is an excellent material for realising thick geometries having relatively high aspect ratios. For this reason, many groups have managed to exploit its chemical and physical properties in microfabrication. The authors of the aforementioned paper have gone one stage further in trying to characterise its electromagnetic properties. To this end, they have managed to successfully measure the refractive index and absorption coefficient, from 100 GHz to 1.6 THz. This is quite difficult to undertake using time-domain spectroscopy, or any other method. The authors then go on to extract values for the dielectric constant and loss tangent, to give meaningful data for future terahertz frequency design. For example, tanδ is quoted to be 6.3 × 10^-6 at 1 THz. This value is very low indeed and the commenter thinks that it would be very tempting to exploit the electromagnetic properties of SU-8 as a low-loss millimetre-wave and submillimetre-wave dielectric. However, using a simple expression to calculate the loss tangent, he shows that at 1 THz, tanδ=0.14. This large value of loss tangent makes the use of SU-8 of little value for low-loss submillimetre wave applications. Moreover, at 100 GHz. it has been found that tanδ=0.08, and is thus of little use as a low-loss millimetre-wave dielectric. The commenter concludes that the work by Arscott et al. is rather misleading and could encourage research groups to invest time, money and resources in an area that will fail to meet expectations