Welding fumes from stainless steel gas metal arc processes contain multiple manganese chemical species

Fumes from a group of gas metal arc welding (GMAW) processes used on stainless steel were generated using three different metal transfer modes and four different shield gases. The objective was to identify and measure manganese (Mn) species in the fumes, and identify processes that are minimal generators of Mn species. The robotic welding system was operated in short-circuit (SC) mode (Ar/CO2 and He/Ar), axial spray (AXS) mode (Ar/O-2 and Ar/CO2), and pulsed axial-spray (PAXS) mode (Ar/O-2). The fumes were analyzed for Mn by a sequential extraction process followed by inductively coupled plasma-atomic emission spectroscopy (ICP-AES) analysis, and by X-ray diffraction (XRD). Total elemental Mn, iron (Fe), chromium (Cr) and nickel (Ni) were separately measured after aqua regia digestion and ICP-AES analysis. Soluble Mn2+, Fe2+, Fe3+, and Ni2+ in a simple biological buffer (phosphate-buffered saline) were determined at pH 7.2 and 5.0 after 2 h incubation at 37 degrees C by ion chromatography. Results indicate that Mn was present in soluble form, acid-soluble form, and acid-soluble form after reduction by hydroxylamine, which represents soluble Mn-0 and Mn2+ compounds, other Mn2+ compounds, and (Mn3+ and Mn4+) compounds, respectively. The dominant fraction was the acid-soluble Mn2+ fraction, but results varied with the process and shield gas. Soluble Mn mass percent in the fume ranged from 0.2 to 0.9%, acid-soluble Mn2+ compounds ranged from 2.6 to 9.3%, and acid plus reducing agent-soluble (Mn3+ and Mn4+) compounds ranged from 0.6 to 5.1%. Total Mn composition ranged from 7 to 15%. XRD results showed fumes had a crystalline content of 90-99% Fe3O4, and showed evidence of multiple Mn oxides, but overlaps and weak signals limited identification. Small amounts of the Mn2+ in the fume (<0.01 to similar to 1% or <0.1 to similar to 10 mg ml(-1)) and Ni2+ (<0.01 to similar to 0.2% or <0.1 to similar to 2 mg ml(-1)) ions were found in biological buffer media, but amounts were highly dependent on pH and the welding process. Mn generation rates for the fractions were tabulated, and the influence of ozone is discussed. The conclusions are that exposures to welding fumes include multiple Mn species, both soluble and insoluble, and that exposures to Mn species vary with specific processes and shield gases.