The influence of waste sample preparation on reproducibility of thermal data

Different milling procedures were applied in waste sample preparation in order to obtain satisfactory reproducibility of thermal investigations (thermogravimetry (TG) and differential scanning calorimetry (DSC)). Due to the complex mixture and the manifold texture involved, municipal solid waste is the most challenging material with regard to obtain representative samples for analysis. The material was subjected to the following milling procedures: ultra-centrifugal mill, cutting mill and vibratory disk mill. Fibrous compounds tend to coil after the milling and sieving process due to hydrogen bonds forming between cellulose materials and electrostatic effects caused by the metallic surface of the sieve. The use of the ultra-centrifugal or cutting mill did not result in an adequate reproducibility. Reproducibility of TG curves, DSC profiles and the effect on resulting enthalpies was evaluated by means of statistical methods. The extensive range of minimum and maximum enthalpies obtained did not allow for correct assessment to be performed. A combination of cutting and vibratory disk mill (agate mill) led to best results. The results are visualized by means of box-plots indicating the minimum and maximum, the 25 and 75% quartile and the median.