Introduction

Thallium is one of the most toxic heavy metals known [1], but the mechanism of its toxicity is still not well understood. Like other heavy metals (e.g., Cd, Pb, Hg), thallium binds to the sulfhydryil groups of proteins and mitochondrial membranes, resulting in the inhibition of a range of enzymes and generalized poisoning. Thallium, in addition, has a high affinity for selenium. Acting as an antagonist of this element [2–4], it interferes with the activities of Se-dependent enzymes, such as glutathione peroxidase, resulting in increased oxidative stress [5]. Possible toxic mechanisms of thallium also include inhibition of cellular respiration, interaction with riboflavin and riboflavin-based cofactors, and distribution of calcium homeostasis. The well-known mechanism of thallium toxicity is related to the interference with the vital potassium-dependent processes, e.g., substitution of potassium in the (Na+/K+)-ATPase. The capability of thallous ions to mimic the biological action of potassium ions has been attributed to the remarkable inability of cell membranes to distinguish between thallium and potassium, possibly due to their similar ionic charges and radii [6].

Although acute thallium intoxications occur infrequently, chronic exposures to even trace amounts of this element may be harmful. The central nervous system is generally recognized as the main target of thallium poisoning [7]. Major symptoms of thallium poisoning include anorexia, headache, and pains in the abdomen, upper arms and thighs, and even in all over the body. In extreme cases, alopecia, blindness, and even death may be caused [8]. The Environmental Protection Agency categorized thallium as a priority pollutant because of its high toxicity [9]. The most anthropogenic sources of thallium are emissions and solid wastes from coal combustion and ferrous and non-ferrous smelting as thallium is a constituent of coal and many sulfide ores [10]. Since thallium compounds are volatile at high temperatures, they are not efficiently retained by electrostatic precipitators or other emission control facilities. Thus, a large fraction of thallium, which enters these processes, is released into the atmosphere [6].

Until the early years of the twentieth century, thallium salts were used extensively to treat venereal diseases (since1883) and to reduce night sweats in tuberculosis patients (since 1898) [11]. The use of thallium salts as rodenticides and later as insecticides began in the 1920s and for the next 45 years remained the principal use for this element [12], thus increasing the likelihood of accidental thallium poisonings in households. Thallium salts have been utilized as a depilatory, being marketed to treat scalp ringworm as they rapidly caused alopecia, enabling ointments to control the fungal infection to be applied more effectively [6].

Nowadays, usage of thallium and its compounds is rather limited. However, there is an increasing demand for thallium in the high-technology and future technology fields [13] since the discovery of high-temperature superconducting components in the system (Tl–Ca–Ba–Cu–O).

Thallium content is obvious to be controlled in environmental and biological objects due to the high toxicity of this element and its behavior as a dangerous anthropogenic pollutant. Thallium levels in biological fluids are units and less micrograms per liter [14]. Mean background thallium concentrations are 0.05–0.5 μg/L for whole blood and 0.25 μg/L for urine. Intoxication level is 8–800 μg/L for blood and >200 μg/L for urine. Lethal concentrations are 4 and 5.2 mg/L for whole blood and urine, respectively [15].

Thus, only sensitive analytical methods could be applied for blood thallium determination below the intoxication level (<8 μg/L), such as inductively coupled plasma mass spectrometry (ICP-MS) [16–20], voltammetry [9, 21, 22], and neutron activation analysis [9, 23, 24]. Atomic absorption spectrometry is a conventional method of trace element determination for clinical purposes. Although there is a trend of graphite furnace atomic absorption spectrometry (GFAAS) gradually yielding to ICP-MS in clinical laboratories [25], it still remains one of the most widespread methods of heavy metal determination in biological and environmental samples in developing countries. However, there are few published articles concerning thallium determination in biological fluids by this method (e.g., [9, 26, 27]). Besides, the obtained limits of detection (50, 80, and 300 μg/L, respectively) appeared insufficient for subtoxic thallium levels.

The aim of the present work was to develop a sensitive technique of blood thallium determination by graphite furnace atomic absorption spectrometry with Zeeman background correction without time-consuming sample digestion or preconcentration analytical stages and to introduce this technique into clinical and forensic practice.

Experimental Section

Apparatus

MGA-915 atomic absorption spectrometer (Lumex, Saint Petersburg, Russia) with high-frequency polarization modulation Zeeman effect background correction system was employed. The principles of correction system performance are presented in [28]. Thallium high-frequency lamp (Lumex) was applied as light source at a wavelength of 276.8 nm, lamp voltage 28 V. Pyrolytically coated graphite tubes with integrated L’vov platforms were used as atomizers. Automated pipettes (altering volume 5–50 μL, Biohit, Helsinki, Finland) were employed for sample introduction.

Akvilon D-301 (Akvilon, Moscow, Russia) water deionization system was employed for water preparation.

Standards and Reagents

All reagents used were of analytical grade and were checked for thallium content before using. The primary standard was prepared by dilution of 1.0 mL of 1 g L−1 Tl aqueous standard solution (Tl+ in 1 M HNO3; Water Investigation and Control Center, Saint Petersburg, Russia, Standard no. 6081-91) to 100 mL with 2 vol.% nitric acid.

Nitric acid solution (2 vol.%) was prepared by the dilution of ultrapure concentrated nitric acid (Reaktiv, Saint Petersburg, Russia) with distilled deionized water (resistance 18.2 MΩ sm).

The chemical modifiers of graphite furnace surface and sample matrix include: hexachloride platinum acid (hexahydrate) H2PtCl6⋅6H2O (Merck, Germany), palladium nitrate Pd(NO3)2 (10 g L−1 solution for atomic absorption analysis, Merck), ruthenium chloride (hydrate) RuCl3⋅xH2O (Merck), rhodium chloride (trihydrate) RhCl3°3H2O (Merck), ammonium nitrate NH4NO3 (Reaktiv), lithium nitrate (trihydrate) LiNO3⋅3H2O (Reaktiv).

Triton X-100 (Amresco, USA) was used for blood sample dilution.

For technique validation, Clincheck® Plasma Control, lyophil, for trace elements (lot no. 417, order no. 8884, Munich, Germany) standard reference material was employed.

Samples

Blood samples were taken from patients of Saint Petersburg toxicological polyclinics using vacuum test tubes (Vacutest®, Vacutest KIMA srl Arzergrande, Italy) with sodium heparinate as anticoagulant (1,400–1,500 U per tube). Blood was taken from the elbow vein.

Procedure

Just before the analysis, blood samples were diluted 1 to 5 per volume with 0.1% Triton X-100. Usage of double-distilled or deionized water is also possible for sample dilution, but it results in worse result precision (RSD is approximately 2% higher than for Triton X-100 solution).

Diluted blood aliquot of 10 μL was introduced into the atomizer with integrated L’vov platform modified with 400 μg Rh. Then, blood matrix modifier (160 μg Pd(NO3)2 and 200 μg NH4NO3) was added into the furnace. Atomic absorption analyses were performed according to the graphite furnace program presented in Table 1.

Table 1 Graphite furnace program optimized for blood thallium determination
Full size table

For thallium quantification, calibration curve method was employed. Calibration graph was plotted using aqueous standard solutions of thallous ions.

Results and Discussion

Analytical Procedure Optimization

Walter Slawin’s stabilized temperature platform concept basic principles [29] were employed in the current work. Thallium is a volatile element, so L’vov platform permanent modification with noble metals (platinum, palladium, ruthenium, rhodium in form of H2PtCl6, Pd(NO3)2, RuCl3, and RhCl3, respectively) was applied. Twenty micrograms of the respective modifier solution was introduced into the atomizer. After that, high-temperature pyrolysis was employed (2,200–2,500°C). Modification procedure was repeated five times in order to achieve total quantity of the corresponding metallic modifier (i.e., Pd, Pt, Ru, or Rh) approximately 400 μg.

Thallium standard solution absorption analytic signal (Tl concentration 10 μg/L) was measured in order to choose the optimum L’vov platform surface modifier.

Data (signal values and RSD) concerning surface modifier optimization are presented in Fig. 1.

Fig. 1
figure 1

Optimization of surface modification for thallium determination (100 pg of Tl, n = 5)

Full size image

As follows from Fig. 1, the best performance is obtained when rhodium is used as the permanent modifier of L’vov platform surface. Both maximum absorption signal and minimum RSD are achieved.

Blood is known as a rather complicated object for trace element analysis. Analytical problems are connected not only with the analytes’ low concentrations but also preliminary with difficult matrix composition, including severe organic matter content and considerable salt background. These factors result in high background absorption and other matrix interferences.

Chloride ion is a common interfering component in GFAAS analysis [30]. Blood contains approximately 0.55 mass% of chloride (or 0.9 mass% of sodium chloride). Graphite furnace program optimization was performed in the presence of chloride ion in a concentration 0.11 mass%, which corresponds to fivefold blood dilution. For pyrolysis and atomization temperature optimization, thallium solution (10 μg/L Tl, 0.11 mass% Cl–) was measured with temperature increment 50°C. Figure 2 illustrates the optimization of pyrolysis temperature as an example.

Fig. 2
figure 2

Optimization of pyrolysis temperature for thallium determination (100 pg of Tl, 0.11 mass% Cl−, n = 5)

Full size image

Optimized graphite furnace program was presented in Table 1.

Unfortunately, the use of permanent platform modification was found insufficient to elude completely chloride interference on thallium determination. Consequently, applying matrix modification is required. Two main ways of chloride interference coping is known:

  1. 1.

    Matrix chloride excess removing in the form of volatile compounds, such as HCl or NH4Cl (NH4Cl sublimation temperature is 338°C), by the addition of sulfuric acid [31] or ammonium nitrate [32].

  2. 2.

    Matrix chloride excess binding into species more stable than TlCl (boiling point ≈725°C, D 0 = 88 kcal/mol), e.g., LiCl (boiling point = 1,380°C, D 0 = 113 kcal/mol) [30].

In the present work, the following matrix modifiers were investigated: sulfuric acid (H2SO4), lithium nitrate (LiNO3), ammonia (NH4OH), and ammonium nitrate (NH4NO3). Two hundred micrograms of the corresponding modifier was introduced into the atomizer. A solution representing fivefold diluted blood was measured in order to choose the best chloride interference coping modifier. In Fig. 3, the performance of matrix modifiers is illustrated.

Fig. 3
figure 3

Choosing of a matrix modifier for thallium determination in chloride containing matrices (10 μg/L of Tl, 0.11 mass% Cl−, n = 5)

Full size image

The addition of 200 μg of NH4NO3 led to thallium absorption signal restoration and provided the most precise results.

High background absorption level is another major problem when determining trace elements in complicated organic-rich matrices by GFAAS. One of the most important conditions for obtaining accurate results when analyzing such samples as whole blood is the use of atomic absorption spectrometer with Zeeman background correction system. However, in some cases, including blood thallium determination, background absorption could be extremely high. Thus, fivefold blood sample dilution with 0.1% Triton X-100 and addition of organic matrix oxidation modifier (160 μg of palladium nitrate) is required.

Analytical Figures of Merit

After the analysis condition optimization had been fulfilled, a calibration graph was plotted using aqueous standard solutions of thallium. The linear range was up to 500 pg. Estimated limit of detection was 0.2 μg/L (3σ criterion, tenfold blank measurement).

Standard reference material (SRM) Clincheck® Plasma Control, lyophil, for trace elements, level II (lot no. 417, order no. 8884, Munich, Germany) analysis was applied for validation. Student’s t criterion was employed for a comparison of the measured value of thallium concentration to SRM’s certified one. The obtained results are presented in Table 2.

Table 2 Technique validation by SRM (Clincheck® plasma control, lyophil, for trace elements, level II) analysis
Full size table

From the data, it follows that the difference between C mean and C certified is insignificant because t calculated < t critical. For validation and precision estimation in the whole dynamic range from 2 to 50 μg/L, spiked blood samples were analyzed. The dependence of the resulting RSD on thallium concentration for spiked blood samples is shown in Fig. 4.

Fig. 4
figure 4

Dependence of RSD (n = 6) on thallium concentration in spiked blood samples

Full size image

As follows from Fig. 4, the precision of the results is in the range 5–11% RSD.

The described technique was applied for analyzing blood samples taken from two rats intentionally exposed to thallium compounds (intraperitoneal injection of thallium(I) nitrate). The obtained results were: 3.4 ± 0.6 μg/L (n = 5, RSD = 11.6%) for rat exposed to 10-mg/kg body weight and 42 ± 5 μg/L (n = 5, RSD = 8.8%) for 100-mg/kg exposure. These results are in agreement with the ones obtained in ANO “Center for Biotic Medicine” (Russia, Moscow) using ICP-MS (spectrometer Elan 9000, PerkinElmer, USA).

Conclusion

Blood thallium determination technique by GFAAS with high-frequency modulation polarization Zeeman background correction without sample digestion had been designed. Analytical procedure was optimized and validated. The described technique was introduced into the practice of Institute of Toxicology (Federal Medico-Biological Agency of Russia, St. Petersburg, Russia).