Fusing Terahertz and Mid Infrared Technologies to Recycle E-waste Black Plastics

Abstract

Plastic is a unique material, it is lightweight, resilient, non-reactive, waterproof, and low-cost. Due to its versatility and use in everyday products and consumption, it has become a huge environmental concern. In 2015, it was reported that 381 million tonnes of plastic waste were produced, black plastics making up 15% and the electronics industry contributing 20% from e-waste. Currently there is no effective sorting technology to sort black or very dark colored plastics as they absorb most of the probing energy. Therefore, they cannot be sorted or recycled and end up in landfills. To solve this problem, a new technique must be established. In this thesis, we propose the use of Terahertz (THz) and Mid-Wave Infrared (MWIR) Spectroscopy with machine learning algorithms to interpret and identify different types of e-waste black plastics. Each black polymer material that interacts with terahertz and MWIR rays affect the shape of the spectrum and has a unique spectral signature. Thus, combining terahertz with MWIR spectroscopy to get multidimensional sensory data will allow us to train a classification model identify black plastics with high accuracy while mitigating the shortcomings of both technologies.