Elemental fingerprinting of gypsum drywall using sector field ICP-MS and multivariate statistics

Abstract

Outgassing of volatile sulphur compounds from gypsum drywall from some sources has resulted in odours, corrosion of wiring and metals and health problems for homeowners. Infrared spectroscopy has been the primary analytical tool to identify 'problematic' drywall. In this paper, we demonstrate that elemental fingerprinting using inductively coupled plasma mass spectrometry and multivariate statistics is an effective alternative. The approach also showed potential in determining the geographic source of gypsum. Nineteen elements (Al, Ba, Ca, Cd, Co, Cr, Cs, Cu, Fe, Mg, Mn, Ni, Pb, Rb, Sr, Tl, U, V and Zn) were measured in 20 samples of drywall, half of which were classified as positive for contamination by Fourier Transform Infrared Spectroscopy. Gypsum from three different mines and a flue gas desulphurisation plant were also analysed. Principal component analysis and multivariate analysis of variance of the elemental data showed significant differences between the problematic and non-problematic drywall and between sources of gypsum. Strontium averaged 1800 ± 500 µg/g in problematic drywall compared to 400 ± 100 µg/g in non-problematic drywall (p < 0.0001).