Geochemical mechanisms of lead stabilization in phosphate-amended urban soils and prediction of physicochemical properties using proximal soil sensors

Abstract

Soil lead (Pb) is one of the most common environmental contaminants in urban areas and remains a persistent global public health concern. Although Pb occurs naturally in soils at average concentrations of approximately 10 mg kg⁻¹, levels in contaminated urban and residential areas frequently exceed 200 mg kg⁻¹, which is the current residential soil limit set by the U.S. Environmental Protection Agency (EPA) to protect human health. However, total soil Pb concentration alone does not adequately represent contamination risk, as soils with similar total Pb concentrations may differ in chemical species that control Pb bioaccessibility. Phosphorus (P) based amendments are widely used to immobilize soil Pb by promoting the formation of relatively stable Pb-phosphate minerals. This process reduces Pb bioaccessibility. Biochar has also gained attention as a complementary remediation material due to its high porosity, large surface area, and ability to modify soil physicochemical properties, which supports this role. However, uncertainties remain regarding the consistency of its stabilization mechanisms and long-term effectiveness under field conditions. The first phase of this doctoral research aimed to elucidate the mechanisms by which organic and inorganic P amendments and biochar reduce Pb bioaccessibility in urban soils. Nine urban soils from Kansas City, Missouri, representing a wide range of Pb contamination levels, were evaluated through 12-month laboratory incubation studies. Four of these studies were followed up by field validation of selected treatments in 1-year-long field experiments. The assessment integrated conventional wet-chemistry analyses with advanced spectroscopic techniques. Bulk X-ray absorption near-edge structure (XANES) was employed for multiple sites for overall changes in soil Pb speciation. In one study, [mu]-X-ray fluorescence ([mu]-XRF) mapping was used to perform spatially resolved Pb speciation using [mu]-XANES to better understand Pb transformation pathways. The results from laboratory incubation and field experiments indicated that bioaccessible Pb was primarily influenced by soil pH and extractable P. Additionally, the P amendments significantly reduced Pb bioaccessibility, from 16% to 36%, and promoted redistribution of Pb into more stable soil fractions. Biochar also reduced desorbed Pb up to 1.67-fold relative to the unamended control. Comparative evaluation showed contrasting immobilization mechanisms among amendments. Mechanisms involving inorganic P treatments were dominated by the formation of pyromorphite-like species, up to 34% according to bulk XANES, while [mu]-XANES identified localized spots with high predominance (> 80%). In contrast, biochar, biochar modified with inorganic P treatment, and organic P treatments primarily immobilized Pb through complexation to organic matter and sorption to Fe-associated phases. These findings demonstrate that inorganic phosphates remain the most effective strategy for promoting the formation of relatively stable Pb minerals with low bioaccessibility. Organic P treatments increased acid phosphatase activity during the early sampling period (1-6 months), whereas inorganic sources generally did not differ from the control. In contrast, [beta]-glucosidase and arylsulfatase had limited sensitivity to the amendments. Enzyme activities appeared to be strongly governed by soil physicochemical properties (pH, TOC, and resin-P) rather than total Pb concentration. These findings indicate that Pb stabilization with P amendments does not necessarily result in uniform recovery of soil biological activity. The second phase of this doctoral research evaluated proximal soil sensing technologies as rapid tools for soil characterization. Although traditional wet-chemistry methods provide accurate measurements, they are limited by low sampling density and high spatial variability, which are typical of agricultural fields. Precision agriculture increasingly relies on proximal soil sensors to generate high-resolution datasets on soil chemical and physical properties that support site-specific management. This study evaluated relationships between multiple soil sensor outputs and selected physicochemical soil properties at two depths and assessed how these influence sensor signal intensity and predictive performance across two contrasting agricultural fields in the U.S. Midwest using the Veris P4000 soil sensor. In this study several soil properties can be estimated with high accuracy when appropriate preprocessing and modeling approaches are applied. For instance, the NIR and VIS sensors had the strongest predictive performance for soil organic carbon and exchangeable phosphorus (R² > 0.7), whereas ECₐ sensor was mainly associated with soil pH, exchangeable cations, and clay content (R² > 0.50). Sensor performance was strongly site-dependent, underscoring the need for site-specific calibration and prior knowledge of local soil conditions to ensure robust interpretation.