Wissenschaftliche Artikel der BAM
Filtern
Dokumenttyp
Sprache
- Englisch (2)
Referierte Publikation
- ja (2)
Schlagworte
- Design of experiments (2) (entfernen)
Organisationseinheit der BAM
This work is aimed at exploring the recovery of heavy metals from the fine fraction of solid waste incineration bottom ash.
For this study, wet-discharged bottom ash fine-fraction samples from full-scale treatment plants in Germany and Sweden were analyzed. The potential for the recovery of heavy metal compounds was investigated through wet density-separation with a shaking table. The feed materials were processed without any pre-treatment and the optimum processing conditions were determined by means of design of experiments. Tilt angle and stroke frequency were identified as the most relevant parameters, and the optimum settings were − 7.5° and 266 rpm, respectively. The obtained balanced copper enrichments (and yields) were 4.4 (41%), 6.2 (28%) and 2.4 (23%). A maximum copper enrichment of 14.5 with 2% yield was achieved, providing a concentrate containing 35.9 wt.% relevant heavy metal elements. This included 26.3 wt.% iron, 4.3 wt.% zinc and 3.8 wt.% copper. In conclusion, density separation with shaking tables can recover heavy metals from bottom ash fine fractions. Medium levels of heavy metal enrichment (e.g., for Cu 2.7–4.4) and yield (Cu: 26–41%) can be reached simultaneously.
However, the separation performance also depends on the individual bottom ash sample.
The demands for the assessment of water quality are increasing steadily, making it necessary to routinely monitor multiple contaminants in water samples. For this application a hapten microarray was developed. In order to reach the required low detection limits a design of experiments (DoE) approach was used to optimize the assay performance. Here we show that a Box-Behnken design plan is an adequate choice for the straightforward exploration of hapten microarray assay parameters. For both read-out systems studied (fluorophore-labelled detection antibodies or enzymatic signal development followed by reflectometric scan, respectively), it was possible to significantly extend the measurement ranges. Furthermore, it could be shown that multivariate data analysis, here partial least squares regression (PLS), can improve the prediction accuracy of 'unknown' samples when used as calibration model, compared to classical, univariate data evaluation methods.