31
samples were taken from the mor layer (
⌀ = 56 mm) and from the 0-10 and 10-
20 cm layers (
⌀ = 32 mm) of the mineral soil (Figure 6). A total of 10 soil
cores were taken along two diagonals over each plot and bulked to one sample.
The vanadium concentrations in the soil samples were analysed by aqua
regia digestion and XANES spectroscopy was applied to the fresh soil samples
to determine the vanadium oxidation state in the bulk solid phase. In addition,
the fresh soil samples were extracted with 0.01 M CaCl
2
(10 g soil:20 ml
solution ratio) to determine the vanadium concentration in the dissolved phase
of the soil and the vanadium oxidation state by means of HPLC-ICP-MS (see
detailed description under “Analytical methods”).
4.4 Analytical methods
4.4.1 X-Ray Absorption Spectroscopy
The XAS measurements were performed using the wiggler beam line 4-3,
Stanford Synchrotron Radiation Lightsource (SSRL, Stanford, USA). Some of
the measurements for collection of the EXAFS spectra were conducted at the
wiggler beamline I811 at MAX-Lab, Lund, Sweden. Both stations were
operated with a Si[111] double crystal monochromator and measurements were
performed in fluorescence mode. The spectra were collected over an energy
ranging from 5235 eV to at least 5645 eV for XANES spectroscopy samples,
covering the vanadium K-edge of 5465 eV (Thompson et al., 2009). For
EXAFS spectroscopy measurements, the energy range was extended up to
6345 eV. The energy was calibrated with a vanadium foil that was measured
simultaneously with, or between, sample measurements.
XANES data analysis
Vanadium K-edge XANES spectra were collected for vanadium sorbed to
ferrihydrite (Paper I), two blast furnace slags (M-kalk and Merit 5000) (Paper
IV) and soil samples collected at the Ringamåla site (Paper V). All spectra
were imported into the Athena software version 08.056 (Ravel & Newville,
2005), where replicate scans were energy-calibrated and subsequently merged.
To enable comparisons between samples, the merged spectra were normalised
as described in Wong et al. (1984). In principle, two parallel lines were fitted
to the pre-edge and post-edge regions of the spectra and the distance between
the lines at E
0
was set to 1.
The pre-edge peak was estimated by means of peak fit analysis in the
Athena programme, in which a baseline together with a combination of Gauss
functions were fitted to the peak. The best fit was selected based on the lowest
R-factor (Equation 4) reported by the programme.
32
?????? − ???????????????????????????????????? =
??????(???????????????????????? − ??????????????????)
2
??????(????????????????????????
2
)
????????????????????????????????????????????????4
The area and maximum intensity were then determined from the net peak
and the pre-edge peak position was established from the centroid position. The
E
1/2
value, which describes the energy at which the normalised intensity equals
0.5, was determined from the normalised spectra.
Vanadium K-edge XANES spectra were also collected for five laboratory
vanadium standards with oxidation states ranging from +3 to +5 (Table 4). The
pre-edge peak features and the position of E
1/2
were evaluated and used as
references to determine the vanadium oxidation states in the other samples.
Different approaches for determining the vanadium oxidation state in unknown
samples have been suggested. A method that employs the positive correlation
between the pre-edge peak intensity and the oxidation state, as applied by
Sutton et al. (2005), was used in Paper V. The standard pre-edge peak intensity
(y) was plotted against the known oxidation state (x), to which a second order
polynomial function was fitted (Equation 5):
?????? = 0.087??????
2
− 0.371?????? + 0.408????????????????????????????????????????????????5
Furthermore, vanadium spectra from sorption experiments conducted on
ferrihydrite (Fh), aluminium (hydr)oxide (HAO) and organic matter (OM)
were included as standards in Paper V. Together with a sample of native
mineral-bound vanadium (inherent V), these standards were used in the LCF
analysis in the Athena programme. The LCF was applied to identify soil
constituents important for vanadium sorption in the Ringamåla soils. The
Table 4. Vanadium XANES standards included in Paper I, IV and V.
Standard
Pre-edge peak
Main edge
Oxidation
Area
Intensity
Centroid position
(eV)
E
1/2
(eV)
state
V
2
O
3(s)
0.39
0.12
5470.2
5476.9
+3
V
2
O
4(s)
1.0
0
0.23
5469.6
5478.4
+4
VO
2+
(aq)
1.0
0
0.36
5469.9
5478.8
+4
V
2
O
5(s)
1.9
0
0.66
5469.4
5480.6
+5
H
2
VO
4
-
(aq)
2.2
0
0.81
5469.8
5481.0
+5
V+Fh (pH4.5)
1.9
0
0.78
5469.9
5481.9
+5
a
V+HAO (pH 6.7)
1.3
0
0.52
5469.8
5481.0
+5
a
V+OM (pH 3.5)
1.0
0
0.36
5469.9
5479.7
+4
a
a
Oxidation state determined based on literature and comparisons with standards.