Preliminary Study for the Determination of Rare Earth Elements Using the ICP-MS Technique

This paper proposes an optimized method for the determination of rare earth elements (Sc, Y, La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu) from soils using ICP-MS technique. First, the soil samples were thermal treated at three different temperatures 550oC, 700oC and 850C in order to eliminate organic matter interferences. Then, the residual samples remaining from the calcination process were extracted in acidic medium with two different digestion methods (method I a mixture of nitric acid and hydrogen peroxide; method II aqua regia mixture) in order to quantify rare earth elements content. The highest recovery percentages for the major rare earth elements analyzed (Sc, Y, La, Ce, Pr, Nd. Sm) were situated in the range 86.13% to 99.90%, in sample residues thermally treated at 700°C and extracted with nitric acid and hydrogen peroxide.

. Locations of soil sampling points phosphate fertilizer. The presence of rare earth elements and metals in soil can affect the quality of food, groundwater, microorganism's activity, plant growth [19,20]. Detection and quantification of metals in soil is usually performed with two sensitive and selective techniques, such as inductively coupled plasma optical emission spectrometry (ICP-OES) and inductively coupled plasma mass spectrometry (ICP-MS). At ultra-trace levels, due to its higher sensitivity, ICP-MS technique is preferred [20][21][22][23][24].
The aim of the study was to establish an optimized procedure for soil pretreatment in order to quantify REEs content at the highest recovery rate, determination of REEs concentration being performed with ICP-MS technique. Thermogravimetric analysis were performed on soil samples in order to establish organic matter content, possible interference in REEs determination. The proposed procedure was applied on agricultural soil samples with clay structure.

2.Materials and methods
Six soil samples were collected from Hunedoara County (P1, P2, P3) and Sibiu County (P4, P5, P6). The sampling points are represented in Figure 1, soil samples with clay structure were collected according to international standards, from 30 cm depth.
The geographic coordinates, description of the sampling points as well as the sample codes are shown in Table 1.

Sample preparation
The soil samples were dried at room temperature and sieved. The fraction with particle dimension less than 150 μm was retained and homogenized prior to analysis.
In order to concentrate the REEs content in soil samples, a thermal treatment was applied at 550ºC, 700ºC, respectively 850ºC for 2h in a calcination oven [24,25].
In Figure 2 is presented the flow treatment of soil samples.

Figure 2. Flow treatment of soil samples
The residues of the samples, obtained after calcination at different temperatures, were digested using the program presented in Table 2. The same program was applied also to the "initial sample" (fig. 2) in order to analyze As, Cd, Cr, Mn, Ni, Pb and Zn.  1  140  100  1600  30  2  140  100  1600  35  3  cooled  cooled  -30 After digestion, the solutions were filtered through a 0.45μm membrane and brought to 25 mL with ultrapure water. Subsequently, the metal concentrations were determined with an ICP-MS equipment.

Materials
All the reagents were of analytical quality grade. Volumetric flask and plastic recipients used for sampling and sample treatments were cleaned with 10% HNO3 ultrapure and then washed with ultrapure water.
Calibration curves for As, Cd, Cr, Mn, Ni, Pb, Zn were performed using 10 mg/L Multielement Certified Reference Material (ICP multi-element standard solution XXI, Certipur, Merck). Quality Based on the obtained results, the REEs were divided in two parts: major elements (scandium, yttrium, lanthanum, cerium, praseodymium, neodymium, samarium) and minor elements (europium, gadolinium, terbium, dysprosium, holmium, erbium, thulium, ytterbium, and lutetium). Two different analytical methods for extracted solutions were developed, one for major elements, situated in the range 20 μg/L to 100 µg/L, and another one for minor elements, situated in the range 10 µg/L to 50 µg/L.
The following performance parameters were evaluated in the experimental test: detection limit (LOD), quantitation limit (LOQ), repeatability (RSDr), intermediate precision (RSDRi), recovery and expanded uncertainty (Uex). The experimental studies applied in order to perform in-house validation for both methods are presented in Table 4. The expanded uncertainty of the analytical results was estimated using the following formulas [26]: where: k is a coverage factor; value 2 for 95% confidence level; combined standard uncertainty; concentration uncertainty (instrument calibration, flasks, pipettes, reference standard material); -25-ml volumetric flask (calibration, temperature); repeatability uncertainty (mass, volume, concentration, extraction recovery); weight uncertainty (balance calibration, linearity); extraction recovery uncertainty. The recovery (%Rec) is defined as: where: is the analyte concentration in the spike sample; is the analyte concentration in the unfortified sample; is the analyte concentration in the added sample.

Thermogravimetric analysis
For structural characterization of the soil samples, a thermogravimetric analysis (TG) was performed, the TG operating parameters being presented in Table 5.

3.Results and discussions
The physical-chemical characterization of soil samples Some specific analysis were performed for the physical-chemical characterization of the soil samples in order to highlight the influence of these characteristics on the tested mixtures extraction capacity for REEs. As is reported in Table 6, a difference between the physical-chemical characterization of soils colected from Hunedoara County and Sibiu County. In this sense, was observed higher conductivity, TOC and total nitrogen were determined at P1, P2 and P3 (sampling sites from Hunedoara County), total nitrogen content being twice than in P4, P5 and P6 soil samples.

Thermogravimetric analyses
Thermogravimetric (TG) analysis was used to describe the decomposition of humic compounds from soil structure, weight transformations as a function of chemical composition being represented at different temperatures. TG analysis is important for obtaining information about the amount of organic matter released in the process and, also about the resulting residue which can be used in the study. Two different samples (P1, P4) were inserted into a crucible of alumina and analyzed. The P1 TG analysis is presented in Figure 3, the mass transformation during the analysis being presented in Table 7. At the end of the process, the residual mass for P1 sample was 85.45 % (Figure 3), the highest content released in the process being correlated with humic mater (Table 7). The TG analysis of P4 sample is presented in Figure 4, the mass transformation during the analysis being presented in Table 8.   For P4 sample, the residual mass remained at the end of the process was 89.61% (figure 4), the highest content released being also humic mater compounds, reported as possible interferences in REEs analyzes [27]. The percentage of mass released in both investigated samples was lower than 15%.

Performance parameters for REEs in-house validation methods
In order to validate the proposed methods, experimental tests were performed on P4 sample, due to high content of REEs (tables 9, 10). The quality control of the results, due to the lack of a commercial certified reference material (CRM) for soil matrix with rare metals, was achieved with a fortified soil sample.
Minor elements were added in the concentrations of 35µg/L while the added solution for major elements was 45µg/L. The recovery percentages were calculated, highest values for major elements were recorded for cerium (99.90%), lanthanum (99.18%), yttrium (96.04%), scandium (92.04%) and samarium (91.55%), Table 9. The mean recovery percentage for minor elements was situated around 86% value (85.80% ± 0.41%), being lower than the one for major elements (93.53% ± 4.12%). Determined values for performance parameters were situated in the accepted range according to literature indication [25].

Metallic elements analyzes
In table 11 are presented the obtained values for As, Cd, Cr, Mn, Ni, Pb and Zn from soil samples (initial sample, figure 2), the results being within the reference values for soil quality according to Romanian Order 756/1997 [28]. To assess the quality control of the analytical results of metallic elements As, Cd, Cr, Mn, Ni, Pb, Zn, a certificate reference material CRM ERM-CC 141 Loam Soil was analysed. The recovery percentage were situated in the range 90.2 to 99.5, as is indicated in Table 12.

REEs analyses
All the soil samples were analysed in duplicate, the results presented in tables 13-15 represent the mean values. The results for REEs major elements (extraction method I) were reported in all analysed soil samples (initial samples, residues at 550°C, 700°C, respectively 800°C), as is presented in table 13.
The results indicate that the content of humic compounds from soil samples influence the extraction of REEs, the samples with highest content of organic compounds having the lowest REEs values (initial sample), and so only approximately 30% of REEs content was extracted. The highest values of REEs were obtained for residues calcinated at 700°C. The explanation of this behaviour probably is represented by the uncomplete decomposition of organic compounds at 550°C (40% to 85% recovery for different elements) and REEs loses at 850°C high temperature (32% to 71% lower recovery).
Highest REEs values were obtained for cerium, lanthanum and scandium in P4, P5 and P6 samples, concentration increase being the following: Ce>La>Sc>Y>Sm> Pr>Nd (Table 13). Regarding REEs minor elements, the results obtained with digestion method I from initial samples, residues at 550°C and 800°C were situated below quantification limits for all investigated elements (Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu). Same observation as in major elements case could be stated, that organic compounds interfere in REEs determination.
In Table 14 are reported only the values of REEs obtained from 700°C residue, the concentrations being either under quantification limits or at very small values, in the range 0.09 mg/kg to 0.27 mg/kg. In samples collected from Hunedoara County (P1-P3), Dy, Ho, Er, Tm, Yb and Lu concentrations were under the LOQ, in contrast, with the soil samples from Sibiu County (P4-P6) were the values were small, but above the LOQ. Table 15 presents the major elements results obtained with digestion method II, extraction of REEs from 700°C residues with aqua regia mixture (usually used method for metal extraction). Reported results are lower than the one obtained with digestion method I. One reason for this behavior could be the interference of chlorine ions with the argon in the ICP-MS technique.
To highlight the differences between the extraction methods results (table 13 and 15), a report of REEs major elements concentration obtained with method II (aqua regia) was divided to REEs major elements concentration resulted from method I (nitric acid and hydrogen peroxide) and plotted in Figure 5. The behavior of soils in aqua regia extraction was different, so it was observed that the selected soils from Hunedoara County have a structure that allows extraction of a larger quantity of yttrium, praseodymium and neodymium. Instead, more quantity of scandium, lanthanum and cerium were extracted with aqua regia from Sibiu County soils. There are no significant differences regarding the samarium concentration.
The results for all minor elements analyzed (Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu), extracted with aqua regia from 700°C residues, were situated below the quantification limits.