Urban airborne lead: X-ray absorption spectroscopy establishes soil as dominant source

Abstract

Background: Despite the dramatic decrease in airborne lead over the past three decades, there are calls for regulatory limits on this potent pediatric neurotoxin lower even than the new (2008) US Environmental Protection Agency standard. To achieve further decreases in airborne lead, what sources would need to be decreased and what costs would ensue? Our aim was to identify and, if possible, quantify the major species (compounds) of lead in recent ambient airborne particulate matter collected in El Paso, TX, USA.

Methodology/principal findings: We used synchrotron-based XAFS (x-ray absorption fine structure) to identify and quantify the major Pb species. XAFS provides molecular-level structural information about a specific element in a bulk sample. Pb-humate is the dominant form of lead in contemporary El Paso air. Pb-humate is a stable, sorbed complex produced exclusively in the humus fraction of Pb-contaminated soils; it also is the major lead species in El Paso soils. Thus such soil must be the dominant source, and its resuspension into the air, the transfer process, providing lead particles to the local air.

Conclusions/significance: Current industrial and commercial activity apparently is not a major source of airborne lead in El Paso, and presumably other locales that have eliminated such traditional sources as leaded gasoline. Instead, local contaminated soil, legacy of earlier anthropogenic Pb releases, serves as a long-term reservoir that gradually leaks particulate lead to the atmosphere. Given the difficulty and expense of large-scale soil remediation or removal, fugitive soil likely constrains a lower limit for airborne lead levels in many urban settings.

Figure 1. Pb-L_III XAFS spectra of PM samples and associated local soils.

Spectra are background-subtracted, normalized, k ²-weighted, and plotted in k-space. Even with repeated and stacked scans, the tiny absolute and relative amount of Pb in the PM yielded noisy XAFS, necessitating acquisition of abbreviated spectra. Because an abbreviated region of k-space was sampled, we present both the XANES and EXAFS region together as a single XAFS fingerprint for comparison of spectra.

Figure 2. Pb-L_III XAFS and derived spectra of model compounds and representative PM and soil.

(A) XAFS spectra. Spectra are background-subtracted, normalized, k ²-weighted, and plotted in k-space. (B) Derivatives of Pb-L_III XANES spectra of model compounds. Derivatives were taken of spectra that had been background-subtracted, normalized, and k ²-weighted. (C) Radial distribution functions (RDF) derived from Fourier transforms of Pb-L_III XAFS spectra of model compounds. Distances are not corrected for phase shift during photo-electron backscattering. Aqueous Pb spectra adapted from Bargar .

Figure 3. Least-squares fits of XAFS spectra of PM with, successively, Pb-humate and Pb-humate+PbSO₄.

The contribution of PbSO₄ to the fit in each of 4 cases was approximately 15%; it did not improve the fit appreciably in two cases and was omitted.