«Dissertation zur Erlangung des Grades “Doktor der Naturwissenschaften” Im Promotionsfach Geowissenschaften Am Fachbereich Chemie, Pharmazie und ...»
The dissolved concentration of Fe and Mn was also measured by flame atomic absorption spectrometry (Varian SpectrAAS 300) to validate the accuracy of the As measurement. The detection limits were 0.5 mg/L for Fe and 0.1 for mg/L for Mn. The recommended values of total Fe in Fe2O3 content are 2.52 ± 0.03 mg/kg (Standard deviation ± 0.10 mg/kg) for the certified reference material (GSR-6, limestone). The actual Fe content in soil samples were evaluated with multiplication of factor 0.699 (molecular weight of Fe/molecular weight of Fe 2O3). It is also noted that for the certified reference material (GSR-6, limestone) the recommended values of Mn content are 434 ± 12 mg/kg (Standard deviation ± 41 mg/kg). The experimental error was defined as the standard deviation of the single measurements that were used for averaging.
Materials and methods The concentration of As, Fe and Mn for the above mentioned six identical agricultural lime samples and reference material were analyzed with X-ray fluorescence spectroscopy (XRF, Philips Bj 2002) to ensure the consistency of the extraction scheme. One sample of the certified reference material (GSR-6, limestone) was prepared and analyzed identically throughout the measurement. For calibration of the XRF 35 certified reference materials (described by Govindaraju, 1989) (Appendix B, table 188.8.131.52 and 184.108.40.206) were used covering As in 1.1 – 400 mgkg-1 range. All the samples had concentrations above the detection limit of the XRF device of As in 1.1 mgkg-1.
All of the above samples were re-analyzed by HG-AAS in the method of standard addition with standard As solution in concentration of 0, 2, 5, 10 and 20 µg/l due to suspect matrix effect in the solution samples. The standard additions technique involves taking the samples, dividing them into separate aliquots, and adding to each increasing quantities of arsenic standard under analysis. These increments are made equal and a minimum of five standard addition mixtures to set up a calibration set. The calibration set consists of one unspiked sample and a set of spiked materials. All of the samples were analyzed and a graph was plotted for analytical absorbance versus the concentration of spike added. These graphs were extrapolated back to the x axis to determine the (negative) concentration of the arsenic presented in the unspiked sample.
Optimum precision of the standard addition increment should be equivalent to the expected concentration of arsenic in the unspiked sample in order to avoid excessive errors in extrapolation from analyzed data. It is also important to ensure the calibration standards posses matrices identical to the samples to be analyzed.
(4.5) Instrumentation of copper The instrumentation parameters of FAAS and XRF could be referred to Appendix C.
(4.6) Sequential extraction procedure of copper Samples were taken from six different specimen bags of limestone powder provided by the producer (Section 4.1). All chemicals were of analytical reagent grade or better (Appendix C). All the collected samples were homogenized, dried and sieved to a powder of grain size 100 µm by Fritsch planetary ball mill in agate made material to assure no metal contamination prior to any chemical treatment. Six subsamples with each of sub-sample from six seperate specimen bags were treated by microwave-assisted total acid (aqua regia) digestion and by a sequential extraction scheme as follows Table (14).
Table (14) Table shows sequential extraction procedure for Copper-bearing solid phases
The pH of the first acetic acid step did not rise beyond a value of 4 (HAc/CaAc buffer) to ensure complete dissolution of the carbonate matrix (Sulkowski and Hirner, 2006). Applied chelating agent to prevent readsorption of copper released at high pH values was not necessary under the above circumstance (Howard and Shu, 1996). Extracting agent targeting organic matter was not required at the absent of this phase in the limestone. The applied extractants in step 2 and 3 showed clear distinction between manganese oxide and iron oxide copper hosts (Tokashiki et al., 2003).
A 1.0g soil sample was added to a 50 ml polypropylene centrifuge tube (Sep-Cor) in addition with a 40 ml volume of the extractant from the step 1 (from top to bottom in sequential order) shown at the above table (14). Soil was extracted with fresh solution listed for each step.
Samples were washed with Milli-Q water (1 μS cm-1) between each different extractant for 30 min. Six subsamples from six different specimen bags of limestone powder provided by the producer, a certified reference material (GBW07108/GSR-6, limestone sample from China) and reagent blank from four different steps shown in table (14) with each of these samples in one duplicate were analyzed.
The applied reagent concentrations and soil-to-extractant ratios were 1:40 (1.0 g to 40 mL) in the first three steps and 1:20 in the last step. An orbital shaker was used for all steps instead of an end-over-end shaker in order to prevent material blow-out due to CO2 overpressure in the first step. The pH was monitored in the second extraction step in order to ensure that no transfer of any CaAc/HAc buffer solution from the first step increased the pH of the second one.
After centrifugation for 30 minutes at 1100 g, the supernatant was decanted into a syringe and filtered through 0.2 μm polycarbonate filters. All extractants were acidified with 1% concentrated HCl immediately after separation and stored in the refrigerator before analysis. The concentrations of dissolved Cu, Mn, and Fe were determined using flame atomic absorption spectrometry (Varian SpectrAAS 300). The sum of all four extraction steps for Cu was at 100% ± 15% (SD of triplicate subsamples out of the 6 specimen bags) in comparison to the amount of Cu determined by the total acid digested sub-samples, the X-ray fluorescence spectroscopy (XRF) analysis and the certified reference material GBW07108/GSR-6 used for analytical quality assessment and control (a commercial limestone CRM from China with an enhanced Cu content of 23 ± 3 mg/kg). The detailed description of Fe and Mn concentration measurements by FAAS could be referred to section 4.4.
Materials and methods (4.7) Site description and extraction procedure of thallium Site description Fig. (14) Location of the collected soil samples and the sampling points (No. 1-20) The above figure 14 showed the collected soil sampling points in pink spots at sample location 1. Table (15) demonstrated the location of each sampling point and the depth of samples collected at each sampling point. The purpose of land use at each sampling points could be referred to Table (19). The sampling location 1 was at the northeast of local cement factory.
Materials and methods
The sampling area covered 45 hectares. Each sampling point represented homogeneously of the local sampling area. The sampling area is broadly divided as land for agricultural purpose and Greenland. The soil sampling core was collected by using stainless steel split-tube sampler (produced by Eijkelkamp company) with no metal contamination sampler was ensured before the sampling process. The size of 5 cm inner diameter of the sampler drill could ensure the top and sub-soils undisturbed during the soil profile removal process. The soil sampling core was immediately transferred to PVC-tubes (produced by Eijkelkamp company) and sample storage fridge before the thallium extraction procedure.
The topsoils sampling points which were identified for agricultural purpose had been subjected extensively to plowing process. On the contrary, the subsoils remained rather intact and closely resemblance with thallium concentration of the parental rocks. The soil core from the agricultural land was characterized with deep dense, dark brown, loamy-sandy till-silty soils. The soil core from the Greenland was characterized with sandy soils.
Fig. (15) Location of the collected soil samples and the sampling points (No. 1-13) The above figure 15 showed the collected soil sampling points in light blue spots at sample location 2. Table (16) demonstrated the location of each sampling point and the depth of samples collected at each sampling point. The purpose of land use at each sampling points could be referred to Table (20). The sampling location 2 was also located at the northeast of cement factory.
The sampling area covered 55 hectares. Each sampling point represented homogeneously of the local sampling area. The sub-sampling points were mainly located in forest with the others being in Greenland and cottage. The soil sampling core was collected by the same method described in above location 1. The soils collected from the sample location 2 were characterized as sandy soils. The sampling points were not involved with agricultural purposes since 1970 and could well be served as Tl concentration reference points to sample location 1.
Extraction procedure of thallium Soil samples were collected from different sampling points described in the above section
4.7. All chemicals were of ultrapure reagent grade (Appendix C). Ultrapure water was used for dilution purpose and the water content was checked following German DIN ISO 11465 prior the experiment. The sample preparation was conducted based on German DIN ISO 11464:1994.
All of the collected soil samples were homogenized, dried at 40 oC in drying oven and sieved to a powder of grain size 2 mm by Fritsch planetary ball mill in agate made material to ensure no metal contamination prior to any chemical treatment. Each soil subsamples listed in above section 4.7 was extracted by ammonium nitrate solution and microwave-assisted total acid (aqua regia) digestion separately. 1M NH4NO3 and aqua regia total acid digestion were used as soil extracts to find out the available and leachable portions of thallium based on operationally defined procedures. NH4NO3 is unbuffered mild extractants which target the exchangeable fraction of the element. The aqua regia total acid digestion aimed to extract total thallium content presented in the sub-samples. The final Tl concentrations were defined in the extractants. The extraction procedure was followed by German DIN 19730:1997-06 and summarized as follows.
All glassware was acid washed to ensure that no metal ions were present on the glass surface. The glassware was first washed in detergent and well rinsed in tap water. The detergentcleaned glassware was rinsed with MilliQ water (~ 18 MΩ cm-1 resistivity) and then heated in 5M HNO3 (Analar) for 4 hours and left for 24 hours before it was rinsed with MilliQ water and heated in a water bath (MilliQ) for a further 4 hours. The clean glassware was again rinsed with MilliQ water and dried in a drying cabinet, before being sealed with clingfilm for storage prior to the usage. The procedure was repeated for polypropylene centrifuge tubes, 50ml polytetrafluoroethylene (PTFE) syringe and PTFE extractant storage bottles.
Materials and methods
A 10.0 ± 0.1 g soil sample was added to a 50 ml polypropylene centrifuge tube (Sep-Cor) after homogenized by the cone and quarter technique and followed by a 25 ml volume of fresh 1 M NH4NO3 solution. The soil solution was transferred to an end-over-end shaker and shaken for 2 hours in 20 revolutions per minute at room temperature (20 ± 2 oC). The soil solution suspension was subsequently centrifugated for 30 minutes at 1100 g and the supernatant was decanted into 50 ml PTFE syringe and filtered through 0.45 μm membrane filters (DIN 19730:1997-06, Section 5.7) with the first 5ml of extractant being discarded and transferred to PTFE storage bottle. All extractants were acidified with 1% concentrated nitric acid (e.g. 0.4 ml nitric acid to 40 ml extractant) immediately after separation and stored in the refrigerator before analysis.
All of the subsamples listed in section 4.7, two certified reference material (GSD-08 and GSS-4) and reagent blank with each of these samples in two duplicate were analyzed immediately after extraction. The applied reagent concentrations and soil-to-extractant ratios were 1:25 (20.0 g to 50 mL) in ammonium nitration extraction and 1:20 in the aqua regia acid digestion. The concentrations of dissolved Tl were determined using optimized inductively coupled plasma mass spectrometry (ICP-MS) (HP 4500 Model) (spike with internal standard) with ten repetitive measurements of Tl counts per second for each sub-sample. An average Tl counts per second against corresponding Tl concentration relationship was deduced after external Tl standard calibration and sub-samples measurements. The amount of Tl in 1 M NH4NO3 extractant solution and total acid digested sub-samples were compared with the certified reference materials GSD-08 and GSS-4 used for analytical quality assessment and control. The recovery rate of Tl for GSD-08 (0.70 ± 0.09 mg/kg) and GSS-4 (0.88 ± 0.15 mg/kg) were shown at 90 and 94% respectively. No solution matrix effect was observed throughout all the ICP-MS measurements.
The geostatistical analyses were performed using the VESPER program (v.1.6). All isoline plots were produced using Surfer 8.0 and GIS software ArcMap (version 9.2). Pair T-test was performed by GraphPad QuickCalcs software. The instrumentation parameters of ICP-MS, could be referred to Appendix C.
(5.0) Results and discussion (5.1) Matrix effect of arsenic Initial HG-AAS measurements showed that the total As concentration (∑5 steps) was not comparable with XRF results (total element content). A suspect matrix effect was discovered on samples in all steps all sub-samples (Appendix B, 9.2.1) under the HG-AAS measurements and the samples were subsequently re-measured by method of standard addition (addition of known amount of As standard) to each sample. The results (figure 16) demonstrated that around “a factor of two“ concentration difference were observed in all samples and all steps when samples were measured by standard addition method in comparison with samples being measured without standard addition..
-5 0 5 10 15 20 25 Arsenic absorbance obtained in HG-AAS against Arsenic concentration in samples