Pavel
Dvořák
a,
Marek
Talába
a,
Jan
Kratzer
b and
Jiří
Dědina
*b
aDepartment of Physical Electronics, Faculty of Science, Masaryk University, Kotlářská 2, Brno 611 37, Czech Republic
bInstitute of Analytical Chemistry of the Czech Academy of Sciences, Veveří 97, 60200 Brno, Czech Republic. E-mail: dedina@biomed.cas.cz
First published on 14th February 2019
In an externally heated quartz atomizer, the most often used hydride atomizer for atomic absorption spectrometry, two-photon absorption laser-induced fluorescence (TALIF) was employed (i) to bring after four decades for the first time conclusive proof of the existence of H radical population sufficient to atomize hydrides thus confirming unambiguously the radical theory of hydride atomization and (ii) to determine the distribution of H radicals in the atomizer. Under typical operating conditions, H radicals are concentrated in an approximately 3 mm long cloud in the center of the optical arm and their peak concentration exceeds 1022 m−3, i.e. four orders of magnitude above the typical analytical concentration of hydride. The lowest detectable H radical concentration is in the order of 1019 m−3. The superb power of TALIF to determine the spatial distribution of H radicals in hydride atomizers for atomic absorption/fluorescence provides a route for elegant optimization of hydride atomization – just by establishing how the atomizer design and parameters influence the distribution of H radicals.
In general, various modifications of quartz tube atomizers (QTAs) and miniature flame atomizers, respectively, are most often employed for AAS and AFS.2 In an ideal case, the only form of analyte present in the observation volume of the atomizer should be free atoms, i.e. there is no decay of free atoms and no atomization interferences. Presently, the optimization of hydride atomization has to be performed by the laborious trial-and-error approach. Obviously, it could be done in a straightforward and elegant way based on the knowledge of what really happens in hydride atomizers. The theory describing what happens in atomizers – the radical theory of hydride atomization2–5 – is based on the formation of free hydrogen atoms (H radicals) at a concentration several orders of magnitude above that predicted by thermodynamic equilibrium calculations.5 The radicals are formed by these reactions:5
H + O2 ⇔ OH + O |
O + H2 ⇔ OH + H |
OH + H2 ⇔ H2O + H |
SeH2 + H → SeH + H2 |
SeH + H → Se + H2 |
Until recently, there was no relevant information on H radical detection in hydride atomizers. Disregarding unpublished attempts to prove the presence of hydrogen atoms in an (evacuated) QTA based on absorption at 121.6 nm and using a catalytic probe,13 the only published evidence on the distribution of H radicals in a miniature flame hydride atomizer was described by Tesfalidet et al.14 who employed electron spin resonance spectroscopy. Unfortunately this approach cannot yield either acceptable spatial resolution or quantitative information on hydrogen atom distribution.
In general, H radicals can be determined in the gaseous phase at atmospheric pressure by spectroscopic methods. A catalytic probe15 can also be used, which is, however, nonselective. Absorption measurements do not provide spatially resolved data directly and they require wavelengths in the vacuum UV range, which makes them challenging. The disadvantage of resonance-enhanced multiphoton ionization16 is the complicated quantitative calibration. Consequently, two-photon absorption laser-induced fluorescence (TALIF) is currently the most suitable method as demonstrated by the determination of atomic hydrogen concentration in discharges17,18 as well as in flames.16,19–21 When measuring H radical density using TALIF, absorption of intense focused laser light with a wavelength of 205 nm is typically employed. The laser excites hydrogen atoms via two-photon absorption from the ground state (1 2S1/2) to the n = 3 states (3 2S1/2, 3 2D3/2 and 3 2D5/2) and the subsequent Hα fluorescence radiation at 656.3 nm is detected.17,18 Since collisional quenching is an important deexcitation mechanism, decreasing the quantum efficiency of the fluorescence, it is necessary to know the quenching rate constants for excited hydrogen atoms. These quenching rate constants for H (n = 3) were published in ref. 18 and 22. The TALIF of H radicals is usually calibrated by measuring TALIF of krypton at known pressure. The cross-sectional ratio for two-photon excitation of krypton and hydrogen was published in ref. 17 and 18 and the ratio of Einstein coefficients for fluorescence emission in ref. 23.
Recently, we presented the first TALIF measurement of H radical concentrations in a plasma discharge ignited at atmospheric pressure, specifically in a dielectric barrier discharge (DBD) atomizer.24–26 However, the hydride atomization mechanism in a DBD atomizer can be, in principle, completely different from that in other hydride atomizers since the processes in the plasma of DBDs are triggered by energetic electrons. The general target of the present work was the most often employed hydride atomizer for AAS – QTA. Our particular aims were (i) to finally bring conclusive proof of the existence of H radical population (sufficient to atomize hydrides) in QTAs and (ii) to determine the distribution of H radicals in the atomizer under typical experimental conditions.
An externally heated quartz atomizer having a rectangular cross-section of the optical (longitudinal) arm with inner dimensions of 7 mm (vertical) × 3 mm (horizontal) and a length of 75 mm was employed (Fig. 2). The optical arm was resistively heated to ca. 850 °C using a wire (not shown in Fig. 2) loosely spiraled around the optical arm to keep the optical arm walls transparent to fluorescence radiation. A quartz tube (2 mm inner diameter, 4 mm outer diameter) was sealed through the center of the bottom horizontal wall of the optical arm to form the vertically oriented inlet arm of the atomizer serving to deliver either Ar with H2 or Kr diluted in Ar. The flow rates of Ar and H2, respectively, were 125 and 15 ml min−1 (H2 flow rate corresponds for example to 1.2 ml min−1 of the delivery rate of 0.5% tetrahydroborate solution27). A silica O2 delivery capillary was centered inside the inlet arm with its tip slid up 1 mm over the junction of the inlet arm with the optical arm.
The laser beam was always (i) parallel to the (longitudinal) axis of the optical arm, (ii) positioned in the middle between the vertical walls of the optical arm and (iii) focused to the center of the optical arm length. Consequently, the axis of the inlet arm always crossed the laser beam in the focus. The following two laser-beam adjustments were used: in the first case designed to maximize the signal-to-noise ratio, a spherical lens was employed to create a circular laser beam cross-section having a diameter in the focus of below 0.1 mm. In the second case designed to realize 2D spatially resolved measurements of the TALIF signal, a cylindrical lens was employed to get a laser beam that was contracted only in the horizontal direction but kept its vertical dimensions. The laser beam cross-section in the focus was ca. 2 mm vertically and below 0.1 mm horizontally. See Fig. S1† for an illustration of the vertical distribution of the relative laser intensity in the focus.
Inside the whole optical arm, the dimensions of the laser beam profile were significantly smaller than the inner dimensions of the optical arm cavity (7 mm × 3 mm). The spatial variations of the laser beam profile were equal for both H detection and Kr calibration. Consequently, the possible decrease of the H TALIF signal outside the focal point of the used lens was reproduced during the calibration measurement and compensated for during the processing of measured data. The fluorescence radiation passing through the vertical wall was recorded perpendicularly using a camera. If not explicitly stated otherwise, the image was corrected to the radiation from the hot resistance wire. The image never included the optical arm sections at a distance higher than ±15 mm from the optical arm centre.
These observations, made under conditions optimum for hydride atomization,2 could seem to disprove the radical theory of hydride atomization in a QTA. However, it should be taken into account that the radical theory predicts the formation of a spatially limited cloud of H radicals at the beginning of the hot zone of the atomizer,2i.e. in the inlet arm 10 to 20 mm upstream from the junction. Consequently, the observations do not contradict the existence of a H radical cloud which is situated in the inlet arm and which vanishes upstream from the junction of the inlet arm with the optical arm. Such a cloud is assumed to fully atomize hydrides. Free analyte atoms formed are thus transported to the optical arm to be detected by AAS.
Since the present experimental arrangement did not make it possible to detect H radicals inside the inlet arm it was necessary to employ in all the following experiments an atomizer setup with an oxygen delivery capillary with its tip slid up 1 mm over the junction of the inlet arm with an optical arm.
Fig. 4 Concentration of H radicals at an O2 flow rate of 1 ml min−1. The interruptions of the shown curve correspond to regions where the TALIF signal was hidden behind the heating wires. |
Fig. 5 shows the distribution of H radicals measured with a laser beam focused with a cylindrical lens which enabled the detection of H radicals in an approximately 2 mm high region above the capillary end. The two dimensional graph shows the central section of the optical arm. Even though the oxygen flow rate here is several times higher than in the previous case (Fig. 3 and 4), the longitudinal size of the H radical cloud is around the same with its center above the inlet arm with the oxygen delivery capillary. The reason for the slight left-hand shift of the cloud is explained above. Because of the higher profile of the laser beam, a higher section of the H radical cloud can be observed (Fig. 5). It can be seen that the radical density decreases with increasing height above the capillary. From the measurements with higher beam positions inside the atomizer optical arm (not shown) it can be concluded that H density becomes negligible in the top sections of the atomizer above the capillary.
The present work is original in terms of bringing proof of the existence of H radical population (sufficient to atomize hydrides) and in terms of determination of the H radical distribution in a QTA.
The above discussed outcome of the radical theory of hydride atomization is that free analyte atoms formed from analyte hydrides are stable within the cloud of H radicals. The most important prerequisite for an ideal hydride atomizer is a complete conversion of the analyte to free atoms and no reactions of free atoms in the observed volume.2 Consequently, the proof that TALIF can be effectively used to determine the spatial distribution of H radicals in hydride atomizers for AAS and AFS opens a way for optimization of hydride atomization in a straightforward and elegant way – just by establishing how the atomizer design and parameters influence the distribution of H radicals. Such an optimization offers potential to positively influence analytical procedures for (ultra)trace element and speciation analysis that are required by research disciplines in environmental, biological and biomedical sciences, and in industry and agriculture.
Footnote |
† Electronic supplementary information (ESI) available: Materials and methods, Fig. S1, notes and references. See DOI: 10.1039/c8sc05655b |
This journal is © The Royal Society of Chemistry 2019 |