Direct Analysis of Environmental Samples by Electrothermal Vaporization into ICP-OES

The standard sample introduction system of inductively coupled plasma (ICP) mass spectrometry (MS) and ICP optical emission spectrometry (OES), which consists of a nebulizer and a spray chamber, has several limitations. It requires that solid samples be in solution for analysis, and the dissolution step can be time-consuming, especially for refractory materials that, by definition, are hard to dissolve. Furthermore, contamination or analyte loss can occur during the multiple steps often involved in the dissolution process. To compound matters, the transport efficiency of conventional nebulization systems is low, with only 2-8% of sample solution typically reaching the ICP, which is why pneumatic nebulization is considered the Achilles' heel of atomic spectroscopy.¹

An alternative sample introduction system that allows direct solid analysis is electrothermal vaporization (ETV). A solid-sampling ETV system is commercially available from Spectral Systems (a German company). A photograph of the system is shown in Figure 1. The outlet of the ETV furnace is connected to the base of the torch using, for example, Teflon tubing (often 1-m long). When equipped with an autosampler, the system allows the automated analysis of up to 50 samples. Typically, a graphite boat contains 1-5 mg of sample, and a pair of automated tweezers inserts each boat into a graphite furnace. Because the furnace acts as a thermochemical reactor, chemical modifiers, such as Cl₂CF₂, CHF₃ or CF₄ gas in the carrier gas, or premixing polytetrafluoroethylene (PTFE) powder with the sample before putting an aliquot in the graphite boat, can be used to transform analytes into more volatile fluorides or chlorides. This decreases the required vaporization temperature, which extends the lifetime of the furnace. A by-pass gas added tangentially at the outlet of the ETV furnace helps to minimize sample loss on the way to the ICP.²

Figure 1. Commercially available ETV system with key components labeled.

Such an ETV system presents several advantages. It enables in situ sample treatment to desolvate it, pyrolyze the matrix and vaporize the analytes. Transport efficiencies approaching 100% can be achieved. Even if some analyte can be lost on the way to the ICP, this translates into a significant improvement in sensitivity compared to nebulization. Furthermore, anything that can be deposited into a graphite boat can be directly analyzed, such as solids, liquids and slurries. Moreover, the direct analysis of solids eliminates the dilution that inevitably results from dissolution, resulting in an even greater improvement in sensitivity compared to nebulization.

However, the introduction of 1-5 mg of sample into the ICP has a visible effect on the plasma, resulting in a significant suppression of Ar emission intensity in ICP-OES during the vaporization step, which is proportional to sample mass.² Point-by-point (time-resolved) internal standardization with an Ar emission line, such as the line at 765.511 nm or 404.442 nm, can thus be carried out to compensate for sample loading effects on the plasma.² This significantly improves the linearity of external calibration curves obtained by introducing an increasing mass of a certified reference material (CRM). Without internal standardization using Ar, sample mass must be limited to 1-2 mg as calibrations are not linear with greater amounts. Following internal standardization, the area of the vaporization peak is integrated and then divided by sample mass as it is very difficult to always introduce the same amount. Coupling ETV to ICP-MS has also been done but solid buildup on the cones in combination with a dry plasma drastically reduces the lifetime of the cones, resulting in the skimmer cone having to be replaced after only 8 h of continuous ETV-ICP-MS operation. No such problem exists with ETV-ICP-OES.

The insertion of a cooling step before the vaporization step was empirically shown to often result in larger vaporization peaks, translating into improved sensitivity and detection limits for several elements in different matrices (rice, soil, clays, etc.), whereas it had no effect for As, Co, Cu, S and Se.^2,3 Analysis of ground rice by ETV-ICP-OES using an external calibration with rice flour NIST 1568a CRM gave similar results to the analysis of digests by ICP-MS (Figure 2), but it required only 77 s per sample, drastically increasing the sample throughput by eliminating the digestion step.³ The lower sensitivity of ICP-OES versus ICP-MS is compensated by the greater sample transport efficiency with ETV by nearly two orders of magnitude versus nebulization, and the elimination of the dilution that is inherent to digestion.

Figure 2. Comparison of elemental concentrations measured in organic white rice directly by ETV-ICP-OES and after acid digestion by ICPMS (n=4-6).³

An example of an ETV-ICP-OES application is geochemical exploration. To identify the location of an undercover ore near Talbot Lake, Manitoba, Canada, till-based soil samples were collected along a line on the surface (Figure 3).² A portion of each sample was digested in aqua regia and analyzed by ICP-MS. The other portion was analyzed directly by ETV-ICP-OES, which readily allows the determination of major, minor and trace elements. Quantitative analysis was done by external calibration with a soil CRM and yielded results in agreement with ICP-MS.² However, because only 1-5 mg of sample is analyzed, the relative standard deviation (RSD) can be as high as 30%, especially for heterogeneous samples such as soils. Grinding and mixing the sample can improve the RSD. Nonetheless, quantification is not necessary for this application, as just comparing the corrected peak areas of the different samples is sufficient to delineate the area where an undercover ore deposit is located (highlighted in yellow in Figure 3). This is a valuable alternative when CRMs are not available. Furthermore, elements like S that are difficult to measure by ICP-MS are readily measured by ICP-OES and can provide further confirmation of where an undercover ore deposit lies.²

Figure 3. Schematic representation (bottom) of an undercover ore deposit; the red dashed line along which samples were collected yielded the ETV-ICPOES profile for P at 178.827 nm (top) reflecting the ore location (yellow highlight).

The Beauchemin group has developed ETV-ICP-OES methods for the simultaneous multielement analysis of a variety of environmental samples (Table 1). In all cases, the plasma was monitored in lateral view mode for maximum robustness. The ETV temperature program depends on the analytes being determined. For example, no F-containing gas modifier can be used when F is being determined. Although F is not ionized in an Ar plasma, precluding its direct determination by ICP-MS, F atomic emission can readily be monitored by ICP-OES. A lower pyrolysis temperature is required for the determination of S, which starts to vaporize during the pyrolysis step, requiring integration of the area under the pyrolysis and vaporization peaks for its quantification.

If pyrolysis of the matrix is complete enough, external calibration with standard solutions can be carried out by adding a desolvation step to the ETV temperature program. This was demonstrated for the accurate determination of 14 rare earth elements in refractory geological materials by ETV-ICP-OES.¹⁴ One condition for accurate analysis was the addition of 50 µL of high-purity water to all solid samples prior to analysis to match the 50 µL aliquots of different standard solutions used for external calibration. This was necessary because water had a visible effect on the plasma beyond the desolvation step, through the pyrolysis and the vaporization steps, and the duration of the increase in plasma brightness was directly proportional to the volume of water. This brighter plasma translated into a significant increase in sensitivity of 58% on average.¹⁴ The ability to calibrate with standard solutions eliminates the constraints linked to solid CRMs, which should match the samples in terms of matrix and analyte concentrations, and are often not available.

Merging the ETV effluent with water vapour generated by infrared heating of the aerosol exiting a Scott double-pass spray chamber, along with a valve between the ETV furnace and the sheathing device used to merge water vapour, which allows venting of the solvent and pyrolysis products, enabled an increase in sample mass up to 13 mg without extinguishing the plasma.¹⁵ This is because water increases plasma robustness by acting as a load buffer in the plasma while constituting a source of hydrogen with higher thermal conductivity than Ar. This resulted in an average improvement in sensitivity by 2.9 ± 1.2 and in detection limits by 5.4 ± 2.7 across 38 emission lines, with similar accuracy but a systematic improvement in RSD from the greater sample mass. An alternative and straightforward way of increasing plasma robustness is to simply add 45 mL/min N₂ or 30 mL/min H₂ as a sheathing gas around the ETV effluent.⁸ Either gas provided more accurate results for the analysis of soils than the water vapour sheath approach.⁸ For the analysis of insects, 45 mL/min N₂ sheathing gas and 30 mL/min H₂ in the carrier gas, after the sample is mixed with PTFE powder in the graphite boat, yielded accurate results for a greater number of elements than N₂ sheathing gas alone.¹¹ Table 1 shows that mixed-gas plasma conditions were beneficial for the analysis of a variety of environmental samples.

Table 1. ETV conditions for quantitative multi-elemental analysis of environmental samples using external calibration with matrix-matched CRM.

Speciation analysis from the solid is even possible with ETV-ICP-OES if different forms of an element have different boiling points. This is illustrated in Figure 4, which shows that different As species vaporized at different times when the ETV temperature program in Table 2 was used. Unfortunately, this could not be used to identify two different As species observed in a soil CRM¹⁶, as As species from aqueous solution vaporized at lower temperatures than the As in the soil CRM. Identifying the two forms remains a work in progress.

Figure 4. Temporal profile obtained for an aqueous mixture of dimethylarsinic acid (DMA), monomethylarsonic acid (MMA), As(III) and As(V) species by ETV-ICP-OES using the ETV temperature program in Table 2.

Table 2. ETV temperature programfor the speciation analysis of As in aqueous solutionby ETV-ICP-OES.

Conclusions

Nonetheless, ETV-ICP-OES allows the analysis of a variety of samples with little (such as washing or grinding) or no sample preparation, which reduces the chances of contamination or analyte loss. However, the ETV and ICP conditions must be optimized for different sample types. The short analysis time of 80-100 s per sample makes the approach ideal for sample screening. Finally, the small amount of sample (1-5 mg) also makes ETV a good match for applications where the sample is precious.

References

(1) Browner, R. F.; Boorn, A. W. Sample introduction: The Achilles' heel of atomic spectroscopy? Anal. Chem. 1984, 56 (7), 786A-798A. https://doi.org/10.1021/ac00271a718

(2) Masquelin, A.-S.; Kaveh, F.; Asfaw, A.; Oates, C.J.; Beauchemin, D. Solid sampling ETV-ICP-OES to study the distribution of elements in clay and soil samples for mineral exploration. Geochem. Explor. Environ. Anal. 2013, 13, 11–20. https://doi.org/10.1144/geochem2012-129

(3) Sadiq, N.; Beauchemin, D. Optimization of the operating conditions of solid sampling electrothermal vaporization coupled to inductively coupled plasma optical emission spectrometry for the sensitive direct analysis of powdered rice. Anal. Chim. Acta 2014, 851, 23–29. https://doi.org/10.1016/j.aca.2014.09.017

(4) Sadiq, N.; Huang, L.; Kaveh, F.; Beauchemin, D. Solid sampling ETV-ICP-OES coupled to a nebulization/pre-evaporation system for direct elemental analysis of glutinous rice by external calibration with standard solutions. Food Chem. 2017, 237, 1–6. https://doi.org/10.1016/j.foodchem.2017.05.063

(5) Maung, P.; Beauchemin, D. Development of a method for the direct determination of fluorine in solid samples using electrothermal vaporization coupled to inductively coupled plasma optical emission spectrometry. J. Anal. At. Spectrom. 2020, 35, 1097–1102. https://doi.org/10.1039/d0ja00079e

(6) Maung, P.; Beauchemin, D. The effect of hydrogen on fluorine detection in solid sampling electrothermal vaporization inductively coupled plasma optical emission spectrometry. J. Anal. At. Spectrom. 2021, 36, 1104–1111. https://doi.org/10.1039/d1ja00090j

(7) Kaveh, F.; Oates, C.J.; Beauchemin, D. Direct analysis of soils by ETV-ICP-AES: A powerful tool for mineral exploration. Geochem. Explor. Environ. Anal. 2014, 14, 305-313. https://doi.org/10.1144/geochem2013-230

(8) Al Hejami, A.; Beauchemin, D. Effect of sheathing the sample aerosol with hydrogen, nitrogen or water vapour on the analytical performance of solid sampling electrothermal vaporisation coupled to inductively coupled plasma optical emission spectrometry. J. Anal. At. Spectrom. 2019, 34, 1426–1432. https://doi.org/10.1039/c8ja00266e

(9) Althobiti, R. A.; Beauchemin, D. Direct multi-element analysis of natural toothbrush by electrothermal vaporization into inductively coupled plasma optical emission spectrometry. J. Anal. At. Spectrom. 2021, 36, 535–539. https://doi.org/10.1039/d0ja00479k

(10) MacConnachie, M.; Lapointe, M.; Galiano, E.; Beauchemin, D. Developing a method for the determination of sulphur and other elements in avian bone and slag using ETV-ICP-OES. J. Anal. At. Spectrom. 2020, 35, 2487–2493. https://doi.org/10.1039/d0ja00288g

(11) Holowaty, Y.; Beauchemin, D. Solid sampling electrothermal vaporization inductively coupled plasma optical emission spectrometry for the analysis of insects. At. Spectrosc. 2024, 45 (3), 226–232. https://doi.org/10.46770/AS.2024.104

(12) Scheffler, G. L.; Sadiq, N.; Pozebon, D.; Beauchemin, D. Risk assessment of trace elements in airborne particulate matter deposited on air filters using solid sampling ETV-ICP-OES to measure total concentrations and leaching with simulated saliva, gastric juice and lung fluid to estimate bioaccessibility. J. Anal. At. Spectrom. 2018, 33, 1486-1492. https://doi.org/10.1039/c8ja00128f

(13) Scheffler, G. L.; Pozebon, D.; Beauchemin, D. Improving the analytical performance of electrothermal vaporization coupled to inductively coupled plasma optical emission spectrometry using a mixed-gas plasma. J. Anal. At. Spectrom. 2019, 34, 891–898. https://doi.org/10.1039/c9ja00010k

(14) Wang, Y.; Kienast, S.; Beauchemin, D. Quick determination of fourteen rare earth elements in powdered refractory materials using electrothermal vaporization with detection by inductively coupled plasma optical emission spectrometry. Anal. Chem. 2025, 97, 19168–19176. https://doi.org/10.1021/acs.analchem.5c02729

(15) Kaveh, F.; Beauchemin, D. Improvement of the capabilities of solid sampling ETV-ICP-OES by coupling ETV to a nebulisation/pre-evaporation system. J. Anal. At. Spectrom. 2014, 29, 1371–1377. https://doi.org/10.1039/c4ja00041b

(16) Beauchemin, D. Environmental analysis by inductively coupled plasma spectrometry: An interesting journey at Queen's University. Can. J. Chem. 2026, 00, 1–10. https://doi.org/10.1139/cjc-2025-0143

About the Author

Diane Beauchemin

With a PhD in 1984 from l’Université de Montréal, Diane Beauchemin is a full professor in Chemistry at Queen’s University. Her research focuses on the fundamentals of ICP-MS and ICP-OES and on expanding their application to environmental analysis, risk assessment of food safety, characterization of nanoparticles, and forensic analysis. In recent years, she received the 2017 Maxxam Award and 2019 Clara Benson Award from the Canadian Society for Chemistry, the 2018 Gerhard Herzberg Award from the Canadian Society for Analytical Sciences and Spectroscopy, and the 2024 Environment Division Research and Development Dima Award from the Chemical Institute of Canada.