Application ICP-OES to Multielement Analysis on Plastic Waste and Blends with Vacuum Gas Oil: Developing a Sample Preparation Protocol

Abstract

This paper introduces a new methodology for a routine metal analysis of plastic waste (PW) and PW blended with petroleum feedstock such as vacuum gas oil and VGO (PW/VGO). For such purposes, recycled polyethylene and polypropylene plastic were selected to mimic the potential feeds to be integrated at the Fluid Catalytic Cracking unit (FCC) to produce valuable products. Elements such as P, Ca, Al, Mg, Na, Zn, B, Fe, Ti, and Si were included in the method development. Different sample preparation methods were evaluated, such as microwave-assisted acid digestion (MWAD) and dry/wet ashing, followed by a fusion of the ash with lithium borate flux. Some PW homogenization pretreatments, such as cryogenic grinding and hot press molding, were also covered. The finding of this work suggests that MWAD with HNO3 and H2O2 is adequate for both types of samples and is the quickest sample preparation; however, the sample needed to be homogenized, and recoveries for Si and Ti may be biased for PW due to the limited solubilities of these elements in the nitric acid media. Carbon removal is required before fusion sample preparation and analysis due to the amount of carbon in PW samples. The sample needed to be homogenized for wet ash fusion but not for the pre-ash (dry) method. A benefit to the damp ash pretreatment is that the ash for the sample was created in the same crucible used for fusion digestion, avoiding material loss during sample management. Fusion from wet ash or carbon removal allowed for better acid solubility for Si and Ti in PW. The results of the PW samples evaluated matched well with those of both sample preparation methodologies. For most elements, precision was <10% regardless of the sample preparation; however, Fe and P had some variation using wet ash fusion, possibly due to contamination in an open digestion system or variation due to being close to the method limit of quantification (LOQ). The methodology reported here is robust enough to be implemented as routine analysis in any laboratory facility.