Investigating the Performance of Different Acid Digestion Methods and Purification for the Rare Earth Elements and Yttrium Determination in Carbonate Rocks.

Abstract

The rare earth elements and yttrium (REY) can act as proxies in geochemical studies to identify the origin and processes involved in the formation of carbonate rocks. This work evaluated critical sample preparation steps for the carbonate analysis and the influence of the matrix on the results.

Introduction

The determination of trace level REY patterns in sedimentary carbonate rocks is a useful approach for understanding depositional conditions of ancient environments. This is due to the lanthanide contraction and the unique tendency of Ce and Eu within the group to undergo oxidation and reduction, respectively. The analysis of REY presents some challenges: matrix effects due to high content of dissolved salts in solution analysis; low concentration of REY in carbonate samples and high content of Ba in sample, which may lead to oxide interferences. These features affect the accuracy and precision of REY determination. In this work, we explore different sample strategies for sample digestion, including steps of chromatographic separation to overcome the effects of high dissolved salts and Ba presence in solution analysis.

Experimental

The acids HNO3 37%, HCl 67%, HF 48%, HClO4 70-72% and CH3COOH (abbreviated as HAc) 99,7% used in this work were all of analytical grade (Merck, Germany), and ultrapure water was produced with an Academic IQ 7000 (Merck, Germany). For this study the reference material JDo-1 (GSJ, Japan) was used for method validation. The samples were collected at Grota da Fervedeira, RN (Silva, 2022), and grinded using a disc mill (Herzog, Germany). The analysis was carried with an ICP-MS 7500ce (Agilent, Japan).

Sample digestion was carried with two methods tested, adapted from in Alexander (2008), and Li et al. (2019). For the whole rock digestion (HF/HClO₄ method)_, 100 mg of sample were weighted in PFA vessels (Savillex, USA). 3 mL of HF and 3 ml of HClO₄ were added with the sample and the placed in a hot plate, at 165°C for 16 h in a closed system, followed by evaporation to dryness at 155 °C. The residual material was redissolved two times in 5 ml of a HCl solution, followed by evaporation to dryness at 120 ºC and redissolution of the resulting residual material in 50 ml of a 0.5 mol/L HNO₃.

For the carbonate phase dissolution (HAc method), 100 mg of samples was weighted samples were transferred to 50 ml polypropylene Falcon tubes containing 30 ml of 1 mol/L HAc. The tubes were kept in an ultrasonic bath at room temperature for 1 h, and an additional aliquot of 20 ml of the 1 mol l^-1 HAc solution was added. The resulting suspension was filtered using cellulose acetate membranes (25 mm, 0.22 µm) and evaporated to dryness at 120 ºC.

After the digestion, three approaches were tested to evaluate the effectiveness of the carbonate analysis. The first one is the diluted sample analysis, in which the redissolved with 50 ml with 0.5 mol/L HNO₃. Samples prepared this way were diluted 10 times with the addition of Indium (In) as the internal standard (IS) prior the analysis by ICP-MS. The second is a chromatographic separation of the REY and the major elements of the matrix, using an AG 50W-X8 cation exchange resin (200-400 mesh, Bio-Rad, USA) After the digestion, 20 mL the extract were evaporated at 90 °C, previously to be dissolved in 2 mL of HCl 1.25 mol/L, and then added to the column. By gravitational pull, 18 mL of HCl 1.25 mol/L elution phase was added, followed by 12 mL of HCl 6 ml/L for REY recovery. The content was then evaporated and dissolved with 10 mL of 0.5 mol/L HNO₃ with the IS prior the analysis. The third, procedure was an additional chromatographic step for Ba removal, using a Ln ion exchange resin (100-150 mesh, Eichrom, USA).This method was applies only in the carbonate samples. After the first chromatographic step, the samples were evaporated again at 90 ºC, followed by successive redissolution with ultrapure water and finally with 2 mL HCl 0.01 mol/L. The sample was added to the column, were 5 mL of HCl 0.01 mol/L were used for an elution phase. After this step, the REYs were recovered for analysis with 15 mL of concentrated HCl solution. The content was then evaporated and dissolved with 10 mL of HNO₃ and the IS prior the analysis.

Results and Discussion

The recovery values of REY for the JDo-1 CRM are presented in Figure 1.

The experimental assays with methods HF/HClO4 and Total HAc provided values of recovery between of 88± 8% (Ce) and 111± 11% (Tm), and 86± 4% (Nd) e 111± 1% (Ho), respectively. The HAc method was then choose to evaluate the recovery of JDo-1 under the purification step. The recovery for the matrix removal yielded values of 93% ±  3 for La and, 110% ±  4 for Ce and 110% ±  5 for Tb.

With the achievement of good accuracy, the samples were analyzed by both Total HAc and HF/HClO4 methods. The need of chromatograph separation of the matrix and Ba depends of the nature of each sample. Sample FERV 12 was not analyzed by the second step chromatographic removal of Ba. The REY normalized pattern of these results is available at Figure 2.

The patterns of Figure 2 reveals that differences on the acid selected for digestion leads to a difference in REY contents of the extract. FERV 12-1 sample difference is observed for the higher content of total REY with the HF/HClO₄ digestion, which are capable of dissolve the silicate phase. The normalized pattern of this sample indicates a small light REE enrichment, small Ce negative and Y positive anomaly. The FERV 12 sample is a pure carbonate samples, but with two different phases of calcite and aragonite. The Total HAc method demonstrated to be ineffective to dissolve all the aragonite phase, providing less REY content in the extract than the HF/HClO₄ method. For this sample, a negative Ce anomaly is observed, in contrast with no exclusive enrichment and small Y anomaly. For the FERV 09 sample, which is composed solely by a calcite, no significative difference of REY amount is noted between the Total HAc and HF/HClO₄ method. For this sample, an enrichment of high REY is observed, with a higher Y positive anomaly.

REY detectability is impaired by the diluted sample analysis, and the matrix removal step is often required for samples with low REY concentration. Furthermore, the dilution may occasionally provoke overcorrections of Eu. The second step of chromatographic separation of Ba can also be a mandatory procedure, but with our results, little difference is observed between the mathematical corrections (Aries, 2000) and the Ba removal step.

Conclusions

This study validated a carbonate analysis method and bought crucial topics of the analytical challenges for the REY determination. The presence of minerals in the matrix affects the REY content, but does not affect the REY normalized pattern. The chromatographic step proved to be a mandatory procedure for improve the detectability. The Eu quantification can be achieved with both correction equations and Ba removal.

Acknowledgements

The authors thanks the CNPq, CAPES and FINATEC for the fellowship and the Federal Police to provide the ICP-MS.