Determination of selenium in biological materials by isotope dilution analysis with an octapole reaction system ICP-MS

Abstract

An inductively coupled plasma quadrupole mass spectrometer (ICP-QMS) with an octapole reaction system (ORS) was used for the determination of selenium by isotope dilution analysis (IDA) in biological reference materials and serum samples from healthy subjects. The potentially interfering argon dimers (⁴⁰Ar³⁸Ar⁺ and ⁴⁰Ar₂⁺) at the selenium masses 78 and 80 were almost eliminated by the use of hydrogen as reaction gas. Thus, the detection limits were improved five times for ⁷⁸Se and the measurement of ⁸⁰Se (the main selenium isotope with an abundance of 49.61%) is now possible. An enriched ⁷⁷Se solution was prepared and characterized by reverse IDA and isotope ratios measured were ⁷⁸Se/⁷⁷Se and ⁸⁰Se/⁷⁷Se. Instrumental parameters were optimised in order to obtain optimal precision and accuracy in the isotope ratio measurement of Se by ORS-ICP-QMS. The precision achieved for the isotope ratio measurements reached 0.2% (RSD for n = 5) for both ratios. Systematic errors, including detector dead time (47 ± 2 ns), mass bias effects (about 3%) and spectroscopic interferences due to the presence of bromine in the samples, were corrected. The accuracy of the measured selenium isotope ratios was improved by correcting the intensity signals of the selenium isotopes for SeH⁺ formation (about 3%). The proposed IDA method has been applied to the determination of Se in biological reference materials (serum, urine and tissues) and the results showed good agreement with the certified values.

---

Introduction

The interest in selenium determination in biological materials has continuously increased because it is considered to be an essential element occurring in inorganic and organic chemical forms.¹ The concentration range in which selenium is physiologically essential (biological window) is narrower than for any other trace element.^2,3 Therefore, the accurate determination of selenium levels in biological materials is required to understand the role of this element in health and disease.

Inductively coupled plasma mass spectrometry (ICP-MS) is a powerful technique for elemental analysis due to its multielemental capabilities, low detection limits and the possibility of measuring isotope ratios.^4,5 In combination with isotope dilution analysis (IDA) highly accurate and precise determinations can be carried out for elements that have at least two isotopes free from spectral interferences.⁶ Since an isotope of the element is used as an “internal standard”, IDA will compensate for many sources of error, including sample loss during the preparation or pre-treatment processes. A major advantage of IDA over conventional calibration procedures is that it can provide compensation for a variety of physical and chemical interferences, such as plasma fluctuations, instrument instabilities and analyte suppressions by sample matrix^7,8,9 This feature is of great importance for the determination of selenium in biological samples where serious matrix effects are commonly observed.

However, the accurate determination of selenium by ICP-MS has been hampered by the presence of spectroscopic interferences. Major isotopes of selenium (⁸⁰Se, ⁷⁸Se and ⁷⁶Se) are all subject to severe Ar₂⁺ interferences. Other Se isotopes of lower abundance (⁸²Se or ⁷⁷Se) can also be interfered by molecular ions from halogens (BrH⁺, ClO⁺) present in the sample. The main polyatomic interferences, which could occur in the determinations of selenium in biological samples, are listed in Table 1.^10–13

Table 1 Spectral interferences on the determination of Se in biological samples by ICP-MS

Se isotope	Abundance (%)	Interference
74	0.89	^38Ar^36Ar^+, ^37Cl2^+, ^40Ar^34S^+
76	9.37	^40Ar^36Ar^+, ^40Ar^36S^+, ^31P2^14N^+
77	7.63	^40Ar^36ArH^+, ^38Ar2H^+, ^40Ar^37Cl^+
78	23.77	^40Ar^38Ar^+, ^31P2^16O^+
80	49.61	^40Ar2^+, ^79BrH^+
82	8.73	^40Ar2H2^+, ^12C^35Cl2^+, ^34S^16O3^+, ^81BrH^+

---

The determination of selenium in biological matrices by IDA-ICP-MS has been reported using a range of sample introduction systems such as pneumatic nebulisation,^14,15 hydride generation^15,16 and, more recently, electrothermal vaporization.^17,18 Nevertheless, these works have reported a series of disadvantages. Thus, Janghorbani et al.¹⁵ reported isotope ratio precisions around 1% using pneumatic nebulisation or hydride generation and measured the less abundant Se isotopes (⁷⁴Se, ⁷⁷Se and ⁸²Se isotopes) in order to avoid Ar₂ interferences. Hydride generation also required intensive sample pretreatment and suffered from reagent impurities and memory effects, as well as from interferences in the hydride generation step. On the other hand, electrothermal vaporization has proved to be effective for the reduction of polyatomic interferences caused from matrix components but not from those caused by polyatomic argon ions. In addition, the precision of such analysis was not as good as those obtained by simple nebulisation.

One way to reduce or eliminate spectral interferences is the use of a sector field mass spectrometer, operated at high resolution.¹⁹ However, drawbacks of the high resolution mass analyser are the inherent decrease in instrument sensitivity as the resolution is increased and its high cost.

The development of ICP-QMS equipped with collision/reaction cells^20–23 is offering an interesting alternative to high-resolution systems for the removal of spectral interferences. The cells (quadrupoles, hexapoles or octapoles) are pressurized with a gas or a mixture of gases to reduce or eliminate the interfering polyatomic species by collisional dissociation and, mainly, by ion–molecule reactions.²⁴ Helium²⁵ is usually used as collision gas while H₂,^20–22 NH₃,²³ Xe²⁵ or CH₄²⁶ are usually employed as possible reaction gases. The collision/reaction cell can also increase the ion transmission efficiencies by collisional focusing.^20,21,27

An analytical methodology for the accurate determination of traces of selenium by IDA-ICP-MS using a new octapole reaction system is described here using hydrogen as the reaction gas. The proposed method has been validated by the determination of selenium in a range of biological reference materials with satisfactory results.

Experimental

Instrumentation

The ICP-MS used was an Agilent, Model 7500c (Agilent Technologies, Tokyo, Japan), consisting of an ICP source with a plasma-shielded torch (grounded metal plate), an enclosed octapole ion guide operated in RF only mode and a quadrupole mass analyser with a secondary electron multiplier operating in a dual mode (i.e., either a pulse counting mode or analogue mode, depending on the ion intensity).

Helium and hydrogen were introduced into the octapole cell as collision or reaction gases, respectively. The gases were introduced into the cell under mass flow control through stainless steel lines. The sample introduction system consisted of a Babington nebuliser with a Scott double-pass quartz spray chamber cooled down to 2 °C. The operating conditions (plasma and reaction/collision cell parameters) and acquisition parameters were optimised for the determination of selenium by IDA and they are summarized on Table 2.

Table 2 ICP mass spectrometer operating conditions and data acquisition parameters

Plasma parameters—
RF power	1500 W
Plasma gas flow rate	15 L min^−1
Auxiliary gas flow rate	1.2 L min^−1
Sample uptake rate	0.4 mL min^−1
Sampling depth	7 mm
Sampler and skimmer	Ni, 1 and 0.4 mm id
Ions lens setting	Optimised for best sensitivity of 10 µg l^−1 Li, Co, Y and Tl, 1% (w/w) HNO3 solution
Reaction/collision parameters—
H2 gas flow	4 mL min^−1
Octapole bias	−13 V
QP bias	−12 V
Data acquisition parameters
Monitored isotopes	76, 77, 78, 79, 80, 81, 82 and 83
Points per peak	3
Acquisition time per point	4 s
Replicates	5

---

A Milestone (Socisole, Italy) Model 1200 microwave digester with an EM-457(A) extractor module and an AC-100 open/close module with medium pressure PTFE vessels were employed for the digestion of the solid samples. The following two steps program was used: (1) 250 W, 2 min and (2) 450 W, 5 min.

Reagents and materials

All reagents used were of analytical grade. Ultra-pure water was obtained from a Milli-Q System (Millipore Co., Bedford, MA, USA). A standard solution of 1000 mg L⁻¹ of Se as SeO₂ stabilized in 2–3% (v/v) nitric acid Suprapur^® was purchased from Merck (Darmstadt, Germany). Hydrogen peroxide (Merck) and nitric acid from Merck (additionally purified by sub-boiling distillation) were used for sample digestions.

Enriched ⁷⁷Se was obtained from Cambridge Isotope Laboratories (Andover, MA, USA) as elemental powder and it was dissolved in a minimum volume of sub-boiled nitric acid and diluted to volume with ultra-pure water. The concentration of this solution was established by reverse isotope dilution analysis. Table 3 gives the isotopic composition of natural Se and the measured isotopic composition of the enriched ⁷⁷Se spike, which was not certified for isotopic composition. The isotopic abundance of the Se standard solution was considered to be of natural isotopic abundance, as reported by Rosman and Taylor.²⁸

Table 3 Isotopic abundances (atom, %) of natural selenium and the ⁷⁷Se enriched spike solution (measured)

Mass	Natural	^77Se enriched^a
a Determined after dead time, SeH^+ formation and mass bias discrimination corrections using natural selenium (n = 5).
74	0.89 ± 0.04	0.063 ± 0.009
76	9.37 ± 0.29	1.03 ± 0.07
77	7.63 ± 0.16	91.1 ± 0.7
78	23.77 ± 0.28	3.9 ± 0.3
80	49.61 ± 0.41	3.4 ± 0.2
82	8.73 ± 0.22	0.54 ± 0.04

---

The biological reference materials analysed in this work were: Seronorm^™ Trace Elements Serum from ″Sero As″ (Billingstad, Norway), Bovine Liver NIST SRM 1577a, Freeze-Dried Urine NIST SRM 2670 from National Institute of Standards and Technology (Gaithersburg, MD, USA) and Horse Kidney IAEA H-8 from the International Atomic Energy Agency (Vienna, Austria). Solid and digested samples were kept refrigerated at 4 °C.

Healthy volunteers were selected from our research group for the human serum samples analysed. Blood samples were extracted from the arterial line using a standard syringe. The blood was allowed to clot and centrifuged at 2000 g for 3 h at 4 °C. All samples were kept refrigerated at 4 °C until measurement.

Procedures

Sample preparation. For Bovine Liver and Horse Kidney reference materials, sample amounts of ca. 0.2 g were weighed directly into polytetrafluoroethylene (PTFE) vessels. An appropriate amount of the ⁷⁷Se spike and 3 ml of HNO₃ (65% w/v) were added. After tightly capping the vessels using the Milestone capping station, the sample carousel was placed in the microwave oven and the heating program started. After cooling, 2 ml of H₂O₂ (30% w/v) were added and the digestion heating program repeated. Finally, the acidified sample solutions were quantitatively transferred into plastic containers and diluted to ca. 20 g with ultra-pure water.

The Seronorm^™ Trace Elements Serum and Freeze Dried Urine reference materials were reconstituted as described in the certificate and directly analysed after appropriate dilution with ultra-pure water. Human serum samples were diluted 1 + 9 with ultra-pure water after the addition of a known amount of the ⁷⁷Se spike solution.

Plastic polypropylene containers were used to prepare all solutions on a weight basis. Diluted standard solutions were prepared daily before the analysis with 1% (w/w) HNO₃.

Analytical procedure for Se determination. For isotope ratio determinations of selenium the isotopes 77, 78 and 80 were selected. The ion lens voltage settings were tuned daily with a 100 ng g⁻¹ Se solution to obtain the maximum signal for these masses. A solution of 1% (w/w) HNO₃ was also used to check the background level caused by polyatomic argon interferences. Typical plasma and ions lens conditions are summarized in Table 2.

The correction for SeH⁺ and BrH⁺ formation was carried out using mathematical equations by monitoring also the signals at masses 76, 82 and 83 (for SeH⁺) and 79 and 81 (for BrH⁺). All isotope ratio measurements were corrected using a dead time of 47 ns. A natural standard of selenium was also measured between the samples to calculate the mass bias correction factor.

The optimum spike to sample ratio was calculated as described previously,²⁹ using the random error propagation theory. Taking into account the natural abundances published by Rosman and Taylor²⁸ and the isotope composition of the spike, the optimal isotope ratios were 0.49 for ⁸⁰Se/⁷⁷Se and 0.36 for ⁷⁸Se/⁷⁷Se.

Finally, the concentration of the analyte was calculated using the isotope dilution equation described previously.²⁹

Results and discussion

Elimination of argon polyatomic interferences

As quadrupole ICP-MS instruments equipped with reaction cells can remove argon-containing polyatomic interferences,²² reaction cell parameters were tuned for optimum signal-to-noise (S/N) ratio by measuring ⁷⁸Se and ⁸⁰Se isotopes. The main instrumental parameter affecting the S/N ratio was the flow of hydrogen used as reaction gas while adequate settings of the octapole bias and QP bias were also important to reduce the background levels by energy discrimination. The optimal cell parameters selected are given in Table 2. When the reaction/collision cell was pressurized, the octapole bias was set at −13 V and the quadrupole bias was at −12 V in order to achieve the lowest background signal. As the electric potential of the quadrupole analyser is slightly more positive than the electric potential of the octapole, the residual gas and product ions generated in the cell have not enough energy to overcome this energy barrier. On the other hand, as polyatomic ions have larger collisional cross sections than those from monoatomic ions, more frequent interactions with the cell gases are produced and a significant reduction in kinetic energy relative to the analyte results.³⁰ Therefore, energy filtering can also be used to help separate the analyte from the polyatomic interference.

Next, the nature and optimum flow of the cell gas were evaluated: firstly, the effects of He, H₂ and He–H₂ mixtures, as reaction and/or collision gases, were tested to remove polyatomic interferences of argon at masses 78 and 80. It is important to recognize that the optimum cell gas flow (minimum limit of detection) does not necessarily correspond to the flow providing maximum net sensitivity (i.e., the detection limit is determined both by the sensitivity and by the noise on the background signal). For ion signals below ca. 10⁴ cps the standard deviation of the background can be approximated by counting statistics. Therefore, the detection limit can be estimated according to:

where EDL is the “estimated detection limit” in ng g⁻¹, B is the background signal (in cps) and S is the net sensitivity in cps per ng g⁻¹. These EDLs can thus be determined as a function of the cell gas flow, as shown in Fig. 1. When the cell was pressurized with He or He–H₂ mixtures, selenium ion signal loss was more rapid than the reduction of argon dimers and no improvement was observed in the EDL. For that reason, H₂ was finally selected as the reaction gas.

Fig. 1 Signal intensity of ⁸⁰Se as a function of the hydrogen flow rate in the octapole reaction cell. (- - -) Signal intensity for the blank solution (1% (w/w) HNO₃). (—) Signal intensity for a 10 ng g⁻¹ selenium standard. ([circle, cut, short horiz bar]) Estimated detection limit.

Fig. 1 shows the effect of H₂ gas flow rate on the signals of both a standard solution of selenium and a blank solution at m/z 80. As can be seen, an important improvement in the EDL was obtained at mass 80. Fig. 1 shows that using a hydrogen flow rate of around 4 mL min⁻¹, the minimum EDL for ⁸⁰Se was obtained (14 pg g⁻¹). The sensitivity for the 10 ng g⁻¹ selenium standard was approximately 50000 counts s⁻¹ at mass 80, while the count rate for the blank (1% (w/w) HNO₃) was 300 counts s⁻¹. This latter value represents a reduction of approximately six orders of magnitude in the intensity of the ⁴⁰Ar₂⁺ dimers compared with conventional ICP-QMS operation. Similar results were observed at m/z 78, where an improvement of five times in the EDL, compared with conventional ICP-QMS, was observed.

Precision of isotope ratio measurements

Acquisition parameters were optimised for the best precision of ⁷⁷Se/⁸⁰Se and ⁷⁷Se/⁷⁸Se isotope ratios using a standard solution of 100 ng g⁻¹ natural Se. Based on our previous experience,³¹ three points per peak and five replicates were selected and only the influence of integration time was evaluated. The influence of the total integration time per mass unit on the relative standard deviation (RSD) (n = 5) of the ⁷⁷Se/⁸⁰Se ratio measurement is showed in Fig. 2. As can been seen, the measured RSD decreased from ca. 1% at 0.1 s integration time per unit mass to ca. 0.2% at 3–6 s total integration time per mass. An integration time of 4 s per mass was finally selected. A similar effect was also observed in the RSD of the ⁷⁷Se/⁷⁸Se ratio measurement.

Fig. 2 Influence of integration time on the isotope ratio precision.

Accuracy of the measured isotope ratios

(a) Detector dead time. For reliable isotope ratio determinations by ICP-MS, the detector dead time has to be appropriately corrected. The detector dead time was determined according to the method proposed by Vanhaecke et al.³² For its evaluation, standard solutions containing 100, 200, 400, 800 and 1200 ng g⁻¹ of Se were prepared and the main isotopes of Se (76, 77, 78, 80 and 82) were measured with the dead time correction “disabled” (that is, set to zero in the software of the instrument) under the instrumental settings shown in Table 2. A dead time value of 47 ± 2 ns was obtained and this value was used to correct the measured intensities in further experiments.

(b) Mass bias. Mass discrimination is another effect causing deviations of measured isotopic ratios from the expected ratios and must be evaluated to produce reliable and accurate results. Correction factors for mass discrimination can be calculated by measuring the isotope ratio of a sample with known isotopic composition: they are then used to correct individual isotope ratios in the samples. Unfortunately, for selenium this is not a simple matter because of the formation of SeH⁺ ions in the octapole, as was observed by Boulyga et al.³³ when using a hexapole cell with hydrogen as reaction gas and by Sloth et al.²⁶ when using a dynamic reaction cell pressurized with methane. They suggested that the SeH⁺ formed could be derived from the sample matrix (i.e. water, methanol, etc.) or from impurities of water in the cell gases.

We observed that, when hydrogen was used to pressurize the octapole cell with a flow rate of 4 ml min⁻¹, the measured ⁸²SeH⁺/⁸²Se⁺ ratio was around 3%, a lower figure than the value around 9% reported by Sloth²⁶ or Boulyga.³³

However, this effect must be taken into consideration because it affects the isotope ratio measurement accuracy by influencing ions of mass m + 1. The SeH⁺/Se⁺ ratio was evaluated with ⁸²Se by measuring the m/z 83/82 ratio in the same run in which the mass discrimination per unit mass was determined. No correction for krypton was necessary as Kr ions were neutralized in the cell. A 100 ng g⁻¹ Se standard of natural isotopic composition was nebulised for this purpose and the measured intensities for ⁷⁷Se and ⁷⁸Se were corrected by taking into account the formation of the hydrides ⁷⁶SeH⁺ and ⁷⁷SeH⁺, respectively. The intensity correction equations used were:

and where f_Se is ratio SeH⁺/Se⁺ based on the measured ⁸³I/⁸²I ratio in the selenium standard solution. Typical values of the selenium factor were around 0.03 (0.028–0.032 depending on the experimental day).

Finally, the mass bias correction was determined using the exponential model of mass bias.³¹Fig. 3 shows the plot of relative error ln(R_exp/R_theo) in the experimental isotope ratios with respect to ⁸⁰Se versus the mass difference between the measured isotopes, ΔM, using weighted linear regression calculations. The result was a linear relationship. The mass bias factor (K), derived from the slope of the regression line was of ca. −3% per mass unit. Similar mass bias factors have been reported for selenium^26,33,34 by ICP-MS.

Fig. 3 Linear relationship between corrected measured isotopic ratios and the mass difference between the measured isotopes from the ⁸⁰Se reference isotope. K = −0.0338.

Once K has been determined, the corrected isotope ratios, R_corr, can be calculated using:

The within-day variation in the mass discrimination factor was also investigated. The temporal stability of the isotope ratio measurement was tested over a period of time of 2 h by measuring a 100 ng g⁻¹ standard solution of Se. As is shown in Fig. 4, no significant drift in the measured isotope ratio (values uncorrected for mass bias) was detected during this period of time. The precision for the 6 individual measurements expressed as RSD was 0.3% for the ⁷⁷Se/⁸⁰Se ratio. This value is close to the precision that can be obtained for each individual isotope ratio measurement. Taking into account the stability of the measured isotope ratios, mass discrimination correction was carried out every 2 h when analysing real samples.

Fig. 4 Stability of ⁷⁷Se/⁸⁰Se isotope ratio with time.

(c) Spectral interferences. For an accurate isotope dilution analysis all the chosen isotopes should be free of significant spectral interferences. The measurement of isotope ratios for unspiked samples provides a simple diagnostic tool to check the presence of such interferences. All possible interferences in the measurement of ⁷⁷Se, ⁷⁸Se and ⁸⁰Se in biological matrices are indicated in Table 1. Chlorine, phosphorus and bromine in the range of concentrations expected in biological samples were studied as possible interferences. No detectable interferences by chlorine at ⁷⁷Se up to a concentration of 400 µg g⁻¹ and by phosphorus at ⁷⁸Se and ⁸⁰Se up to a concentration of 400 ng g⁻¹ were observed. However, the interference of ⁷⁹BrH⁺ on ⁸⁰Se was important. Fig. 5 shows the effect of increasing amounts of bromine (as Br⁻) on the ⁷⁷Se/⁸⁰Se isotope ratio (solid line). The presence of 1 µg g⁻¹ of Br caused an error of about 56% in the ⁷⁷Se/⁸⁰Se ratio in a solution of 20 ng g⁻¹ of Se. Janghorbani et al.¹⁵ also reported systematic bias in the ratio ⁸²Se/⁷⁷Se caused by ⁸¹BrH⁺ interferences in the determination of selenium in biological matrices. The uncertainties indicated in Fig. 5 for the uncorrected ratios correspond to the standard deviation of the measurements.

Fig. 5 ⁷⁷Se/⁸⁰Se isotope ratio as a function of bromine concentration. (—) Original data. (····) Corrected ⁷⁷Se/⁸⁰Se isotope ratio after proposed mathematical correction.

This interference was corrected by monitoring also ⁷⁹Br and ⁸¹Br (% abundance 50.69 and 49.31, respectively²⁸) by considering equations (1) and (2) and the following equations for masses 79, 80, 81 and 82:

and where f_Br is the ratio BrH⁺/Br⁺. For the calculation of the bromine factor, f_Br, the measured intensities at mass 82 for the different concentrations of bromine were plotted against the measured intensities at mass 81 after subtracting the contribution of selenium at both masses (⁸²Se⁺ and ⁸⁰SeH⁺), using a selenium standard solution in the absence of bromine, and combining equations (5) and (6). The slope of the line gives the bromine factor directly. For the experiment shown in Fig. 5, the bromine factor was 0.087 ± 0.004. Once the bromine factor was calculated, the intensity of selenium at mass 80 was calculated using equations (1), (2), (3) and (4) and using the known value for the selenium factor (0.03).

The corrected ⁷⁷Se/⁸⁰Se isotope ratios are plotted in Fig. 5 as a function of bromine concentration (broken line). As can be observed, the correction method used is adequate within the experimental uncertainty. It is worth mentioning here that, when the concentration of bromine increases, so does the uncertainty in the correction at mass 80. This fact was taken into account in the calculation of the uncertainties shown in Fig. 5, where the law of propagation of errors was applied to equations (1) to (6) and to the corrected 77/80 ratio.

Validation of the method and determination of selenium in real samples

Selenium was determined in four biological reference materials and in three human serum samples by the proposed IDA procedure. In all cases, the samples were spiked with the enriched isotope, digested (or simply diluted) and the ratios ⁸⁰Se/⁷⁷Se and ⁷⁸Se/⁷⁷Se evaluated after adequate interference correction. As an example, the obtained mass spectra in the range 76–83 for a serum and a blank are shown in Fig. 6. The presence of high levels of bromine is clearly seen from the spectrum. As similar levels of bromine were also detected in the reference materials, the correction of BrH⁺ had to be performed in all samples for the ⁸⁰Se/⁷⁷Se ratio. This was carried out by measuring a 1 ppm bromine standard and calculating the bromine factor from the measured 82/81 and 80/79 ratios. A bromine factor of 0.0415 was applied for the correction of all intensities in the real samples and reference materials. This factor was found to change slightly from day to day. The selenium factor was calculated using the mass bias correction solution (100 ppb natural selenium standard) from the measured 83/82 ratio. Then, all intensities measured were corrected using equations [1] to [6]. The final quantitative results using both ratios are given in Table 4. As can be observed, the results obtained for all reference materials for the 78/77 ratio lie within the ranges given for the certificate at the 95% confidence level. However, for Horse Kidney and Bovine Liver data for the 80/77 ratio show certain differences from the certified values, which could be a result of the low concentration of selenium in the digested samples and the high value of the bromine correction.

Fig. 6 Mass spectrum of a serum sample diluted 1 + 9 (solid line) and an ultra pure water blank (broken line).

Table 4 Analysis of biological certified reference materials

Material	Reference value	Concentration found^c
		^80Se/^77Se	^78Se/^77Se
a Concentrations in ng g^−1. b Concentrations in µg g^−1. c Average ± uncertainty at 95% confidence level.
Seronorm^™ Human Serum^a	83 ± 3	83.6 ± 1.7	83.2 ± 2.0
SRM-2670 Freeze Dried Urine^a	30 ± 8	31.1 ± 1.2	30.3 ± 0.5
IAEA H-8 Horse Kidney^b	4.67 ± 0.30	5.3 ± 0.5	4.59 ± 0.07
NIST SRM 1577a Bovine Liver^b	0.71 ± 0.07	0.87 ± 0.05	0.70 ± 0.07

---

Real human serum samples were also analysed to compare our results with literature values and as a starting point to carry out an analytical survey of the levels of selenium in the population of Asturias (Spain). Results from 3 individuals are given in Table 5. As can be observed, the levels found are within the ranges given in the literature.³⁵ The values obtained for the 80/77 ratio are somehow higher than those obtained for the 78/77 ratio, indicating the need for a better correction procedure for the interference of bromine on selenium. In the future this method will be applied to study selenium level changes in health and disease.

Table 5 Analysis of selenium in human serum samples from healthy subjects by IDA-ICP-QMS with an octapole reaction system (uncertainly expressed 95% confidence level)

Sample	Concentration found/ng g^−1
	^80Se/^77Se	^78Se/^77Se
1	77.8 ± 1.4	73.4 ± 2.2
2	97.6 ± 2.4	89.6 ± 2.0
3	87.3 ± 0.6	83.0 ± 0.6

---

Acknowledgements

The authors are grateful to Agilent Technologies for the loan of the Agilent 7500c to the Ministry of Science and Technology (Madrid, Spain), for financial support through project number BQU2000-0221 and to CONACYT (Mexico) for awarding a Doctoral Grant to L. Hinojosa Reyes.

Healthy volunteers (V. Diaz, A. Rodríguez, L. Bravo, C. García and G. Centineo) are greatly acknowledged for blood donation and M. Serrano of the “Hospital Central de Asturias” for the blood extraction.

References

1. L. H. Foster and S. Sumar, Crit. Rev. Food Sci. Nutr., 1997, 37(3), 211.
2. L. A. Daniels, Biol. Trace Elem. Res., 1996, 54, 185.
3. M. Navarro-Alarcón and M. C. López-Martínez, Sci. Total Environ., 2000, 249, 347.
4. M. Stastná, I. Nemcova and J. Zýka, Anal. Lett., 1999, 32(13), 2531.
5. C. Sariego Muñiz, J. M. Marchante Gayón, J. I. García Alonso and A. Sanz-Medel, J. Anal. At. Spectrom., 1999, 14, 193.
6. K. G. Heumann, Int. J. Mass. Spectrom. Ion Processes, 1992, 118(119), 575.
7. F. Vanhaecke, L. Moens and R. Dams, J. Anal. At. Spectrom., 1998, 13, 1189.
8. J. P. Valles Mota, M. R. Fernández de la Campa, J. I. García Alonso and A. Sanz-Medel, J. Anal. At. Spectrom., 1999, 14, 113.
9. C. Sariego-Muñiz, J. M. Marchante Gayón, J. I. García Alonso and A. Sanz-Medel, J. Anal. At. Spectrom., 1999, 14, 1505.
10. J. Goossens, F. Vanhaecke, L. Moens and R. Dams, Anal. Chim. Acta, 1993, 283, 137.
11. S. H. Tan and G. Horlick, Appl. Spectrosc., 1986, 4, 445.
12. L. S. Zhang and S. M. Combs, J. Anal. At. Spectrom., 1996, 11, 1049.
13. F. H. Konad and M. H. Yang, J. Anal. At. Spectrom., 1996, 11, 413.
14. B. T. G. Ting, C. S. Mooers and M. Janghorbani, Analyst, 1989, 114, 667.
15. M. Janghorbani and B. T. G. Ting, Anal. Chem., 1989, 61, 701.
16. W. T. Buckley, J. J. Budac and D. V. Godfrey, Anal. Chem., 1992, 64, 724.
17. J. W. H. Lam, R. E. Sturgeon and J. W. McLaren, Spectrochim. Acta, Part B, 1999, 54, 443.
18. J. Turner, S. J. Hill, E. H. Evans, B. Fairman and C. S. J. Wolff Briche, J. Anal. At. Spectrom., 2000, 15, 743.
19. N. Jakubowski, L. Moens and F. Vanhaecke, Spectrochim. Acta, Part B, 1998, 53, 1739.
20. I. Feldmann, N. Jakubowski and D. Stuewer, Fresenius’ J. Anal. Chem., 1999, 365, 415.
21. I. Feldmann, N. Jakubowski and D. Stuewer, Fresenius’ J. Anal. Chem., 1999, 365, 422.
22. P. Leonhard, R. Pepelnik, A. Prange, N. Yamada and T. Yamada, J. Anal. At. Spectrom., 2002, 17, 189.
23. V. I. Baranov and S. D. Tanner, J. Anal. At. Spectrom., 1999, 14, 1133.
24. S. D. Tanner, V. I. Baranov and D. R. Bandura, Spectrochim. Acta, Part B, 2002, 57, 1361.
25. J. T. Rowan and R. S. Houk, Appl. Spectrosc., 1989, 43, 976.
26. J. J. Sloth and E. H. Larsen, J. Anal. At. Spectrom., 2000, 15, 669.
27. D. R. Bandura, V. I. Baranov and S. D. Tanner, Fresenius’ J. Anal. Chem., 2001, 370, 454.
28. K. R. Rosman and P. D. P. Taylor, J. Anal. At. Spectrom., 1999, 14, 5N.
29. J. I. García Alonso, Anal. Chim. Acta, 1995, 312, 57.
30. N. Yamada, J. Takahashi and K. Sakata, J. Anal. At. Spectrom., 2002, 17, 1213.
31. J. Ruiz Encinar, J. I. García Alonso, A. Sanz-Medel, S. Main and P. J. Turner, J. Anal. At. Spectrom., 2001, 16, 315.
32. F. Vanhaecke, G. Wannemacker, L. Moens, R. Dams, C. Latkoczy, T. Prohaska and G. Stingeder, J. Anal. At. Spectrom., 1998, 13, 567.
33. S. F. Boulyga and J. S. Becker, Fresenius’ J. Anal. Chem., 2001, 370, 618.
34. S. M. Gallus and K. G. Heumann, J. Anal. At. Spectrom., 1996, 11, 887.
35. S. Caroli, A. Alimonti, E. Coni, F. Petrucci, O. Senofonte and N. Violante, CRC Crit. Rev. Anal. Chem., 1994, 24, 363.

---