Читать книгу X-Ray Fluorescence in Biological Sciences - Группа авторов - Страница 33

As aforementioned, one of the main advantages of XRF in comparing with other atomic spectroscopic techniques is the potential to obtain multielemental information of vegetation samples with a relatively simple sample treatment. At present, there are different sample treatment strategies to be used in combination with XRF techniques [29]. The choice of the most adequate one depends on the purpose of the analysis, the amount of sample available as well as the requirements inherent to each XRF configuration. In Figure 2.3, a summary of the main sample treatment procedures used in vegetation analysis by XRF is displayed.

Figure 2.3 Most commonly used sample treatment strategies for vegetation sample analysis using XRF techniques.

One of the most widespread sample preparation strategies for vegetation bulk‐mass sample analysis when using WDXRF, 2D‐EDXRF, and 3D‐EDXRF systems is the fabrication of pressed pellets from the powdered plant material. Prior to the grinding of the sample into fine powder, vegetation specimens are usually freeze dried or oven dried at T ≤ 60° to remove water content of the matrix. Then, to reduce the vegetal material to a fine powder a milling procedure is employed using a grinder (i.e. ball mill). In this sense it is of significance to choose a suitable container (i.e. agate, silicon carbide, boron carbide, tungsten carbide…) in the grinding process in order to overcome contamination problems arising from the grinding material. This fact is of special relevance when elements at trace levels should be determined. For plant sample analysis by XRF, grain size effects are relatively small in comparison with other environmental solid samples such as soils. In fact, a study performed by Omote and co‐workers showed that when the grain size is ≤710 μm, the measured X‐ray intensity became constant [30]. In view of the morphology of plant powder and the capacity to be compacted together, the fabrication of pressed pellets can be performed without the addition of a binder [29]. Nevertheless, in some applications a binding agent (free of contaminant elements and with a low X‐ray absorption coefficient) is added before pelletizing to make pressed pellets more stable and to elude distortion of the pellet's flat surface over time. The most commonly used binding agnets include boric acid, cellulose and wax, and the amount of these substances to be used depends strongly on the vegetal matrix, the amount of sample used to prepare the pellet, and the pellet size. For instance, to make solid pellets of 3.2 cm in diameter, it was necessary to add 150–200 mg of cellulose to 500 mg of savannah grass powder [31]. To press the pellets usually requires a hydraulic or manual press and an adequate die set, which includes a die body, base, a plunger and two polished metal disks. These disks are usually made of a metal alloy such as tungsten carbide which proved to be useful for pressing this type of samples. At present, there are a broad variety of analytical methods in the literature to prepare vegetation pressed pellets involving different sample amounts (usually 0.5–5 g) and pellet dimensions (10–40 mm).

If the amount of plant sample available is limited (i.e. roots), the option to make a stable pellet is unpractical and in this case the vegetation sample can be presented to the XRF spectrometer as a loose powder by packing it in a cell or by distributing the powder on an adhesive tape or X‐ray foil (made of polyester, polypropylene, polycarbonate, etc.). Using this latter approach, a plant layer ranging from 0.8–2.5 mg/cm² can be obtained [32]. However, in view of the inherent X‐ray absorption by the foil, the sensitivity for low Z elements such as Na is reduced in comparison with the analysis of pressed pellets. Another shortcoming of loose powder preparation is the worse precision of the results obtained due to the bulk density and grain size differences between replicate measurements. In any case, depending on the analytical purpose, the loose powder preparation method can be a fast analytical alternative above all when the amount of sample is limited and WDXRF, 2D‐EDXRF, and 3D‐EDXRF systems are used.

Another interesting alternative when the amount of vegetation sample to be analyzed is scarce is the use of TXRF systems. As it was discussed in the previous section, to perform analysis by TXRF the sample should be presented as a thin film by deposition of a small amount of sample (a few μg or μl) on a reflective carrier. In terms of sample preparation procedure, an additional advantage of TXRF is the possibility to prepare easily the vegetation sample by suspending several milligrams of the powdered sample (10–50 mg) in an adequate disperser and depositing a few μl (5–20 μl) of the suspension on a reflective carrier. This approach has demonstrated to be effective, for example, in multi‐elemental analysis of vegetal foodstuff [24]. Nevertheless, it is interesting to remark that a careful study of all the parameters affecting suspension preparation and TXRF analysis (i.e. amount of sample, suspension concentration, dispersant type, sample deposition volume on the reflector, etc.) should be evaluated in order to obtain reliable results. Another important aspect to be considered to prepare homogeneous solid suspensions and to minimize the particle size agglomerations caused by the presence of particle superficial electrostatic forces when using suspension preparation, is to grind the sample to a particle size <100 μm and use high energy ultrasound treatments before sample deposition on the reflector. Finally, TXRF analysis is especially suited when calibration standards with a similar matrix to the analyzed samples are not available, since quantification can be performed by means of internal standardization. Using this approach, acceptable results in terms of accuracy and precision were obtained for the determination of mid‐high Z elements (Mn‐Sr) in vegetal foodstuff. But empirical calibration using several plant reference materials was necessary to correct absorption effects and obtain reliable results for low Z elements such as K and Ca.

Another more sophisticated sample treatment strategy to be used in combination with TXRF to improve limit of detection for trace elements in vegetation samples is digestion. Despite the fact that this approach implies longer analysis times and the consumption of reagents, the use of TXRF as a detection technique still entails lower costs than ICP based techniques as no consumables or cooling media are necessary in most of benchtop TXRF systems. Usually plant samples can be decomposed, as other biological matrices, using a mixture of nitric acid and hydrogen peroxide [33, 34]. However, as mentioned above, the silicon content in some plant species is quite high (up to 10%) and the use of hydrofluoric acid is needed to obtain a complete dissolution of the sample [3]. Classically, to perform the digestion procedure, digestions at atmospheric pressure using heat sources such as sand baths and hot plates are used. However, at present, microwave ovens are widely used as a faster alternative. Additional advantages of microwave digestions include the limiting of reagents consumption and also the decrease of analyte losses and sample contamination. Block heaters for digestion of small amounts of vegetation samples (several mg) have also been successfully employed [24]. Despite the fact that suspension and digestion are the most commonly used sample preparation strategies when analyzing vegetation samples by TXRF, other novel treatments have been recently developed. For instance, in a recent publication, tree leaves of different species were prepared by a novel device that first compressed the leaves between two 75 mm thick organic foils and then cut out a 30 mm diameter disc which was then stuck to a reflector and inserted into the TXRF system. Results were compared with those from microwave digestion and it was concluded that the non‐destructive direct method allowed rapid screening of leave samples with acceptable detection limits (0.5–18 mg/kg) [35].

Finally, the direct analysis of the plant material is necessary in some applications, above all when dealing with the study element distribution within plant tissues or in in vivo analysis. In this latter case, the specimen is directly analyzed by the XRF system and any sample treatment is not required. For example, the study of the intake of mineral nutrients by time‐resolved XRF analysis proved to be useful in the investigation of plant diseases due to nutrient deficiency and excess [36].

Usually, when performing μ‐XRF analysis to study element distribution within plant tissues a minimum sample treatment is necessary to ensure a flat and smooth surface of the vegetation sample which can affect the quality and reliability of μ‐XRF spectra collected. It has been demonstrated in different studies that dehydration of plant tissues can alter the in‐situ location of dissolved components [25]. For this reason, vegetation tissues are usually sandwiched between two thin films transparent to X‐rays to prevent the sample from drying, oxidizing and dehydrating [37]. More sophisticated sample treatments for μ‐XRF analysis include sectioning of the plant tissue. Different approaches are used for such purposes including the sectioning of the fresh tissue under cryogenic conditions or after embedding the tissue in a hard resin or soft paraffin wax. In the latter case, it is possible to obtain ultrathin (100 nm to 1 μm) sections that are very well‐suited for elemental distribution and speciation analysis of plant tissues using synchrotron radiation [38].

It is worth to mention that in p‐EDXRF systems vegetation samples can be measured in‐situ in the field without any sample treatment directly by pointing the handheld window (of approximately 2 cm²) to the vegetal tissue. Even under field‐moist conditions, it has been demonstrated, for instance, that p‐EDXRF systems can be reasonably used for the determination of Zn and Cu in different vegetation species [18]. However, a significant improvement of the results can be assessed if drying the vegetation sample or if using additional sample treatment such as the fabrication of pressed pellets from the powdered plant material [19].