Simultaneous quantification of 49 elements associated to e-waste in human blood by ICP-MS for routine analysis

Graphical abstract


Background information
The intensive use of consumer electronics generates hundreds of thousands of tons of highly polluting waste (electronic waste) all over the world [1]. In the manufacture of these devices, many rare earth elements and minor minerals are used, and as a consequence those elements are appearing as emerging pollutants in many areas of the planet [2]. Pollution is particularly important in developing countries where the processing of electronic waste is done informally and openly on many occasions [3]. There is increasing evidence that these elements can harm the health of humans and other vertebrates [4], and it is becoming increasingly necessary to include these rare elements in biomonitoring protocols [5,6]. Currently, there are few ICP-MS validated methodologies which allows to simultaneously analyze heavy metals and rare earth elements in human blood. For this reason, in this work, a fast, simple, and robust method has been developed and validated for the determination of the main elements associated with electronic waste in human blood.

Instrumentation
An Agilent 7900 ICP-MS (Agilent Technologies, Tokyo, Japan) equipped with standard nickel cones, Ultra High Matrix Introduction (UHMI) system and a cross-flow nebulizer with a Make Up Gas Port (X400 Nebulizer, Savillex Corporation, MN, USA) was used for all measurements. UHMI maximizes the plasma robustness of the 7900 ICP-MS through a combination of aerosol dilution and automated optimization of plasma temperature.
The 4th generation Octopole Reaction System (ORS) was able to measure all elements in helium (He) mode, even though, the low-mass elements which are normally measured in no gas mode due to the lack of interferences [7]. The high sensitivity of the 7900 ICP-MS allowed that all measurements were performed in He mode to maximize sample throughput, as was the case in this study.
The instrument parameters are described in Table 1. The full data was recorded with Agilent MassHunter Data Acquisition software (version 4.2) and processed with Agilent MassHunter Data Analysis software (version 4.2).

Procedures Solutions
The alkaline solution were prepared adding 1

Calibration curves
Two calibration curves from 0 ng mL À1 to 300 ng mL À1 were prepared using the multi-element mixtures A and B described above with thirty-one elements and nineteen elements respectively [8]. The internal standard, a mixture of Ge, Ir, Rh and Sc, were added to all calibration points and samples to give a final concentration of 37 ng mL À1 . Sample preparation 0.130 mL of sample was diluted by a factor of ten using the alkaline solution [9]. Then, the internal standard was added and the mixture was shaken. The blanks consisted of 0.130 mL of alkaline solution and the same amount of the internal standard was added.

Recovery studies
Recovery studies were performed without any blood matrix. They were performed in alkaline solution due to the commercial lack of a whole blood free of the target elements. For this purpose, the alkaline solution was fortified at a levels of 0.1, 0.5 and 5 ng mL À1 and at level of 0.01, 0.1 ng mL À1 using the mixtures A and B, respectively. Then, the recovery values were calculated dividing the obtained concentration by the theoretical concentration calculated from the added amount of the solution standard A and B. Three independent replicates of each concentration were prepared, and each one of them analyzed by triplicate.

Limit of quantification (LOQ)
The LOQ of the proposed methodology were calculated by quantifying the blank values in ten independent replicates of 0.130 mL of alkaline solution following the sample preparation protocol describe above. For this purpose, the corresponding amount of internal standard was added in all blanks. The LOQ were calculated as three times the concentration of the blanks for all elements.

Calibration curves and limit of quantification (LOQ)
Two calibration curves were performed in order to avoid the interferences of doubly charge ions from some rare earth elements, such the case of Nd and Sm (double charge ions with an m/z ratio from 71 to 75) which interfere with Ge and As. For the same reason, Ge was removed as internal standard Table 2 Limit of Quantification in ng mL À1 (LOQ) obtained for ten independent replicates. LOQ has been calculated as three times the concentration of the blank solution. The LOQ for 0.130 mL of sample volume was found to be lower than 0.4 ng mL À1 and 0.06 ng mL À1 for toxic heavy elements and rare earth elements, respectively ( Table 2). Only for Ti, Fe, Cu, Zn Sr, Sn and Ga higher LOQ were observed ( Table 2). According to these results the proposed methodology can be efficiently applied for the quantification of trace elements in whole blood for biomonitoring analysis, and the obtained results can be fully compared to those reported in the bibliography [10,11].

Accuracy and precision
The accuracy and precision of the proposed methodology was assessed performing recovery studies using alkaline solution fortified at three different levels of concentration, as previously described. The sample preparation was emulated by diluting with the same alkaline solution in the same way described above for the samples (1:10 v/v).
Figs. 1 and 2 show the recoveries obtained, which ranged from 89 to 128% for REE, and from 87 to 118% for toxic heavy elements. In general, the calculated relative standard deviations (RSD) were lower than 8%. However, for some elements (Ti, Cr, Cu, Ni, Se, Fe, Ba, Zn, Sm), the RSD raised to 15-16% at the at the lowest level of fortification. On the other hand, the precision was improved at the highest level of concentration studies, as it was lower than 5% for all elements.

Assessment of the methodology
The proposed methodology was assessed by analyzing aliquots of a human blood material with a known concentration of several elements (ClinCal whole blood calibrator, RECIPE Chemicals + Instruments GmbH, Munich, Germany). Two dilution factors, 1:10 and 1:20, were carried out in order to study the effect of the blood matrix. Three independent replicates were taken from the same blood vial and three blood vials were also analyzed. All samples were analyzed by a triplicate each one. The results obtained are given in Table 3. As it can be seen, the accuracy for As and Se was improved when a higher dilution was selected (1:20 vs. 1:10). The reason behind this may be the ionization of these compounds could be affect by the whole blood matrix. However, as shown in Table 3 the accuracy for the rest 47 elements was good at both dilution levels, and the calculated concentrations were in agreement with the certified values for all elements reported. The RSD values (%) obtained for three independent replicates of the material were lower than 5% except for As, Hg and Se. Our results show that this method is valid for quantification of these elements in human whole blood samples without the need of additional clean-up steps, and without appreciable matrix effects for all elements.