Abstract
High-resolution continuum source molecular absorption of the calcium mono-fluoride molecule CaF in a graphite furnace has been used to determine fluorine in tea after acid digestion, alkaline solubilization and preparation of a conventional aqueous infusion. The strongest absorption ‘line’ of the CaF molecule is at 606.440 nm, which is part of the rotational fine structure of the X2Σ+ − A2Π electronic transition; it has a bond dissociation energy of 529 kJ mol−1, which is comparable with other molecules used for fluorine determination. One advantage of using Ca as the molecule-forming reagent is that spectral interferences are extremely unlikely in the spectral range of its strongest absorption. Another advantage is that Ca acts both as molecule forming reagent and chemical modifier, so that no other reagent has to be added, making the method very simple. The only disadvantage is that Ca has a somewhat negative influence on the graphite tube lifetime. The limit of detection was found to be 0.16 mg L−1 F, corresponding to 1.6 ng F absolute, and the calibration curve was linear in the range between 0.5 and 25 mg L−1 with a correlation coefficient of R = 0.9994. The results obtained for a certified tea reference material were in agreement with the certified value on a 95% confidence level. There was also no difference between the results obtained after an acid digestion and an alkaline solubilization for 10 tea samples, based on a paired t-test. The values found in the 10 samples ranged between 42 μg g−1 and 87 μg g−1 F; the tea infusions contained between 21 μg g−1 and 56 μg g−1 F, with an extraction rate between 48% and 74%.
Highlights
► Fluorine is determined using high-resolution molecular absorption spectrometry. ► The absorption of the CaF molecule in a graphite furnace was investigated. ► Fluorine is determined in several tea samples.
Keywords
High-resolution continuum source molecular absorption spectrometry; Fluorine determination; CaF molecular absorption; Electrothermal vaporization; Tea analysis
1. Introduction
Fluorine and several of its compounds are extremely toxic, but the fluoride ion is also an essential element for humans. Its beneficial action in the prevention of dental caries was discussed controversially for many years, but is generally accepted nowadays. Because of its importance for human health, fluorides are for example added to drinking water and toothpaste; however, too much fluoride can also result in fluorosis. For these reasons fluoride has to be determined routinely in a variety of samples [1], [2], [3] and [4].
Tea is one of the most important non-alcoholic beverages, and about two third of the world's population consume it daily in the morning [5], and many of them also throughout the day. The plants used to make tea are well known to accumulate fluoride [6], [7] and [8], and the fluorine content might even be used as a parameter for judging the tea quality [9]. Moderate consumption of tea has several health benefits, and one of them is the attenuation of dental caries due to the presence of fluoride [10]. Excessive consumption of tea with high levels of fluoride, however, increases the risk of dental fluorosis, particularly in countries where the water is already enriched with this element [11] and [12]. The occurrence of this disease has already been documented in countries like China, where tea is a very popular beverage, and often consumed all day long [13] and [14].
The determination of fluoride is widely described in the literature and is well-reviewed for example by Huang et al. [15]. The methods range from classical gravimetric and volumetric methods over photometric [1] and electrochemical methods, such as fluoride ion selective electrode (ISE) [4], to ion chromatography (IC) [16], [17] and [18]. Predominating methods nowadays are ISE and IC; the ISE is easy to use, cost-effective and suitable for continuous monitoring, whereas IC is more expensive and requires better trained personnel. Common for both methods is the response for only free fluoride ions; organic or covalently bonded F cannot be detected, making the application practicable only for aqueous systems. Ions such as Al3+, Si4+, Mn3+, Mn2+, and Fe3+ or water-soluble organic compounds form complexes with, and/or absorb free fluoride ions, causing serious interferences resulting in low recoveries.
Fluorine is the most electronegative element with a very high ionization potential of 17.42 eV, and its resonance lines are located in the vacuum-UV range below 100 nm. Hence, the use of spectrometric methods such as inductively coupled plasma optical emission spectrometry (ICP OES) or atomic absorption spectrometry (AAS) is not feasible for fluorine determination. Similarly, inductively coupled plasma mass spectrometry (ICP-MS) is also not capable to determine fluorine, as the ionization potential of the argon plasma is not sufficient to ionize this element, unless electrothermal vaporization is used for ‘dry’ sample introduction [19] and [20].
Molecular absorption spectrometry (MAS) in flames and furnaces was proposed as an alternative for fluorine determination already in the 1970s, as summarized in recent Review articles [21] and [22]. Particularly Dittrich [23] and [24] and Tsunoda et al. [25] investigated molecular absorption spectra of gallium, indium and aluminum mono-halides in graphite tube furnaces, and aluminum mono-fluoride was applied in most of the later applications for quantitative fluorine determination using MAS. Dittrich [23] and [24] used a dual-channel spectrometer and a hydrogen hollow cathode lamp (HCL) for his investigations; in all the other studies conventional line source AAS was used with a platinum HCL, as one of the lines emitted by the latter coincided accidentally with the AlF spectrum, and D2 background correction. Electron excitation spectra of diatomic molecules such as AlF, which are found in the UV and visible range of the spectrum, are characterized by a pronounced rotational fine structure [21], [22] and [26]. The spectrometers normally used for classical AAS are not capable of resolving this fine structure, resulting in spectral interferences and background correction errors when applied for MAS, so that this approach never found general acceptance.
In the last decade, however, the development and commercial availability of equipment for high-resolution continuum source AAS (HR-CS AAS) [26] has shed new light on the determination of non-metals using MAS [21] and [22], and the determination of fluorine using HR-CS MAS was described by Huang et al. [15] for flame MAS, and by Heitmann et al. [27] using a graphite furnace. The main advantage using this new technique is indeed the high resolution provided by the double monochromator with a prism for pre-dispersion and an echelle grating for highest resolution [26] that almost completely resolves the fine structure of the molecular bands, avoiding at the same time spectral interference and background correction errors [21] and [22]. In both of the above papers the GaF molecular absorption band head at 211.248 nm was chosen instead of the AlF spectrum because of its narrower profile and greater freedom from spectral interference [15] and [27]. Gleisner et al. [28] optimized the procedure, making it more robust using various additives and modifiers; however, at the expense of a more complex procedure, and a relatively high blank value due to the large quantities of reagents that had to be added. Later, Gleisner et al. [29] applied the procedure for the determination of fluorine in toothpaste.
The goal of this work has been to develop a simple and reliable method for the determination of fluorine in tea using MAS and a commercially available high-resolution continuum source atomic absorption spectrometer with electrothermal vaporization for the generation of the target molecule. At the same time we were trying to simplify the method described by Gleisner et al. [28], searching for other molecule-forming agents and avoiding the large number of reagents that were necessary for this procedure, as the proposed method was intended for routine application.
2. Experimental part
2.1. Instrumentation
A high-resolution continuum source atomic absorption spectrometer Model contrAA 700 (Analytik Jena AG, Jena, Germany) with a flame and a transversely heated graphite tube atomizer in two separate sample compartments was used for all measurements. The instrument is equipped with a 300 W xenon short-arc lamp, operating in a hot-spot mode, as continuous radiation source for the wavelength range from 185 to 900 nm, a high-resolution double monochromator, consisting of a prism pre-monochromator and an echelle grating monochromator, providing a spectral bandwidth per pixel of about 2 pm at 200 nm, and a linear charge coupled device (CCD) array detector with 588 pixels, 200 of which are used for analytical purposes, displaying the vicinity of the analytical line at high resolution.
All measurements were performed using the graphite furnace technique and pyrolytically coated graphite tubes with integrated PIN platform (Analytik Jena Part No. 407-A81.025) and a sample volume of 10 μL. The integrated absorbance of three pixels has been added (peak volume selected absorbance, PVSA, AΣ3,int) [30], resulting in the best signal-to-noise ratio. The temperature program used for vaporization and determination of fluorine via the CaF molecular absorption at 606.440 nm is shown in Table 1.
Table 1.
Temperature program used for the determination of fluorine via the molecular absorption of CaF using HR-CS ET MAS; argon purge gas flow rate 2.0 L min−1 in all stages except during vaporization, where the gas flow was interrupted.
Stage Temperature (°C) Ramp (°C s−1) Hold (s)
Drying 1 80 6 10
Drying 2 110 5 10
Pyrolysis 725 300 10
Vaporization 2250 3000 5
Cleaning 2650 500 4
Table options