Search

Artisanal gold mining can lead to soil contamination with potentially toxic elements (PTEs), necessitating soil quality monitoring due to environmental and human health risks. However, determining PTE levels through acid digestion is time-consuming, generates chemical waste, and requires significant resources. As an alternative, portable X-ray fluorescence (pXRF) offers a faster, more cost-effective, and sustainable analysis. This study compared total As, Ba, Cr, Cu, Fe, Mn, Ni, Pb, Sr, Ti, V, and Zn obtained from pXRF with their pseudo-total contents obtained through acid digestion (USEPA method 3051A) in areas influenced by artisanal gold mining in the Eastern Amazon, Brazil. pXRF data and machine learning algorithms were used to predict extractable Cu, Fe, Mn, and Zn. Linear regression models were fitted to compare the two methods, and random forest and support vector machine techniques were used to predict extractable contents. The best regression model fits for the pseudo-total PTE contents were those for Cu, Fe, Mn and Pb in agricultural areas (R 2  > 0.80); Fe and Mn in gold mining residue (R 2  > 0.70); and Ba, Cu and Mn in urban areas (R 2  > 0.80). The best models for predicting the extractable PTE contents were those for Cu (R 2  = 0.72; RMSE = 2.58 mg dm -3 ) and Zn (R 2  = 0.71; RMSE = 1.44 mg dm -3 ) in agricultural areas and for Zn (R 2  = 0.72; RMSE = 0.43 mg dm -3 ) in gold mining residue. The results demonstrated that pXRF can characterize and predict PTE contents in mining-impacted areas, offering a sustainable approach to soil quality analysis., (© 2024. The Author(s), under exclusive licence to Springer Nature B.V.)

Soil quality monitoring in mining rehabilitation areas is a crucial step to validate the effectiveness of the adopted recovery strategy, especially in critical areas for environmental conservation, such as the Brazilian Amazon. The use of portable X-ray fluorescence (pXRF) spectrometry allows a rapid quantification of several soil chemical elements, with low cost and without residue generation, being an alternative for clean and accurate environmental monitoring. Thus, this work aimed to assess soil quality in mining areas with different stages of environmental rehabilitation based on predictions of soil fertility properties through pXRF along with four machine learning algorithms (projection pursuit regression, PPR; support vector machine, SVM; cubist regression, CR; and random forest, RF) in the Eastern Brazilian Amazon. Sandstone and iron mines in different chronological stages of rehabilitation (initial, intermediate, and advanced) were evaluated, in addition to non-rehabilitated and native forest areas. A total of 81 soil samples (26 from sandstone mine and 55 from iron mine) were analyzed by both traditional wet-chemistry methods and pXRF. The available/exchangeable contents of K, Ca, B, Fe, and Al, in addition to H+Al, cation exchange capacity at pH = 7, Al saturation, soil organic matter, pH, sum of bases, base saturation, clay, and sand were accurately predicted (R 2 > 0.70) using pXRF data, with emphasis on the prediction of Fe (R 2 = 0.93), clay content (R 2 = 0.81), H+Al (R 2 = 0.81), and K + (R 2 = 0.85). The best predictive models were developed by RF and CR (86%) and when considering pXRF data + mining area + stage of rehabilitation (73%). The results highlight the potential of pXRF to accurately assess soil properties in environmental rehabilitation areas in the Amazon region (yet scarcely evaluated under this approach), promoting a more agile and cheaper preliminary diagnosis compared to traditional methods., (© 2024. The Author(s), under exclusive licence to Springer Nature Switzerland AG.)

Farms use large quantities of fertilizers from many sources, making quality control a challenging task, as the traditional wet-chemistry analyses are expensive, time consuming and not environmentally-friendly. As an alternative, this work proposes the use of portable X-ray fluorescence (pXRF) spectrometry and machine learning algorithms for rapid and low-cost estimation of macro and micronutrient contents in mineral and organic fertilizers. Four machine learning algorithms were tested. Whole (i.e., as delivered by the manufacturer) (CP) and ground (AQ) samples (429 in total) were analyzed to test the effect of fertilizer granulometry in prediction performance. Model validation indicated highly accurate predictions of macro (N: R 2  = 0.92; P: 0.97; K: 0.99; Ca: 0.94, Mg: 0.98; S: 0.96) and micronutrients (B: 0.99; Cu: 0.99; Fe: 0.98; Mn: 0.91; Zn: 0.94) for both organic and mineral fertilizers. RPD values ranged from 2.31 to 9.23 for AQ samples, and Random Forest and Cubist Regression were the algorithms with the best performances. Even samples analyzed as they were received from the manufacturer (i.e., no grinding) provided accurate predictions, which accelerate the confirmation of nutrient contents contained in fertilizers. Results demonstrated the potential of pXRF data coupled with machine learning algorithms to assess nutrient composition in both mineral and organic fertilizers with high accuracy, allowing for clean, fast and accurate quality control. Sensor-driven quality assessment of fertilizers improves soil and plant health, crop management efficiency and food security with a reduced environmental footprint., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2023 Elsevier Inc. All rights reserved.)

Ca and Mg are the most important chemical elements in lime. Properly measuring Ca and Mg contents is essential to assess the quality of lime products. Quality control guarantees the adequate use of lime in industrial processes, in soils, and helps avoiding adulteration. Proximal sensors can aid in this process by determining Ca and Mg contents easily, rapidly and without producing chemical waste. The objective of this study was to evaluate the use an environmentally-friendly method of analyzing the quality of lime. We studied 1) the use of portable X-ray fluorescence (pXRF) to predict concentrations of Ca and Mg in lime, 2) tested if NixPro™ sensor can improve prediction accuracy and 3) tested if sample preparation methods (grinding) affect analyses. 74 samples of lime were analyzed by two different laboratories (lab. 1 = 38, lab. 2 = 36). All samples submitted to pXRF and NixPro™ analyses. Sensor analyses were done in whole (CP) and ground (AQ) samples to test the effect of sample preparation in prediction performance. High correlation was found between Ca and Mg contents measured via pXRF and laboratory analyses. Mg-CP presented the highest correlation coefficient (r = 0.81); Mg-AQ, the lowest (0.57). Predictions presented good performance (R 2  > 0.68); Mg had the best results (0.86). Separating models per laboratory showed that some datasets are harder to model, probably due to variability in the source material (limestone). The addition of NixPro™ data contributed to improve prediction accuracy, although slightly. Predictions using CP samples presented the best results, especially for Mg, indicating that grinding is not necessary. This pioneer study demonstrated that fused proximal sensors can be used to rapidly and easily determine contents of Ca and Mg in soil amendments without producing chemical waste., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper, (Copyright © 2023 Elsevier Inc. All rights reserved.)

Several materials have been characterized using proximal sensors, but still incipient efforts have been driven to plant tissues. Eucalyptus spp. cultivation in Brazil covers approximately 7.47 million hectares, requiring faster methods to assess plant nutritional status. This study applies portable X-ray fluorescence (pXRF) spectrometry to (i) distinguish Eucalyptus clones using pre-processed pXRF data; and (ii) predict the contents of eleven nutrients in the leaves of Eucalyptus (B, Ca, Cu, Fe, K, Mg, Mn, N, P, S, and Zn) aiming to accelerate the diagnosis of nutrient deficiency. Nine hundred and twenty samples of Eucalyptus leaves were collected, oven-dried, ground, and analyzed using acid-digestion (conventional method) and using pXRF. Six machine learning algorithms were trained with 70% of pXRF data to model conventional results and the remaining 30% were used to validate the models using root mean square error (RMSE) and coefficient of determination (R 2 ). The principal component analysis clearly distinguished developmental stages based on pXRF data. Nine nutrients were accurately predicted, including N (not detected using pXRF spectrometry). Results for B and Mg were less satisfactory. This method can substantially accelerate decision-making and reduce costs for Eucalyptus foliar analysis, constituting an ecofriendly approach which should be tested for other crops.