We report on experiments that evaluate the potential of terahertz (THz) time-domain spectroscopy (TDS) for quality control of polymeric compounds. We investigate specimens out of a polyethylene compound with silver-coated titanium dioxide nanospheres and a glass-fiber reinforced epoxy composite. We further examine an industrial polymer product produced by injection molding. Our data demonstrates that THz imaging is a powerful tool for contactless quality control in the polymer industry.

No abstract

The authors present terahertz time-domain spectroscopy (TDS) measurements on injection molded liquid crystalline polymers (LCPs). Pure LCP yields a strong birefringence of Δn=0.2. They find that the direction of the optical axis varies within the injection molded sample. By terahertz (polarization) imaging domains of certain directions of the optical axis could be observed. These strongly depend on the mold geometry. Since the alignment of particles and rodlike polymer molecules tend to follow the melt flow during injection molding, terahertz TDS can be used to analyze injection molded LCP parts and control the process performance.

The macroscopic and the microscopic diffusion coefficients of a phospholipid spin label (16-PC) in the model membrane 1-palmitoyl-2-oleoyl-sn-glycero-phosphatidylcholine have been measured simultaneously in the same sample utilizing the new technique of spectral-spatial electron spin resonance imaging. The macroscopic diffusion coefficient Dmacro for self-diffusion of 16-PC spin label is obtained from imaging the concentration profiles as a function of time, and it is (2.3 +/- 0.4) x 10(-8) cm2/s at 22 degrees C. The microscopic diffusion coefficient Dmicro for relative diffusion of the spin probes is obtained from the variation of the spectral line broadening with spin label concentration, which is due to spin-spin interactions. Dmicro is found to be substantially greater than Dmacro for the same sample at the same conditions, and is estimated to be at least (1.0 +/- 0.4) x 10(-7) cm2/s. Possible sources for their difference are briefly discussed in terms of the models used for Dmicro.