An unconventional method for measuring the Tc L₃-edge of technetium compounds

Tc L₃-edge XANES spectra have been collected on powder samples of SrTcO₃ (octahedral Tc⁴⁺) and NH₄TcO₄ (tetrahedral Tc⁷⁺) immobilized in an epoxy resin. Features in the Tc L₃-edge XANES spectra are compared with the pre-edge feature of the Tc K-edge as well as other 4d transition metal L₃-edges. Evidence of crystal field splitting is obvious in the Tc L₃-edge, which is sensitive to the coordination number and oxidation state of the Tc cation. The Tc L₃ absorption edge energy difference between SrTcO₃ (Tc⁴⁺) and NH₄TcO₄ (Tc⁷⁺) shows that the energy shift at the Tc L₃-edge is an effective tool for studying changes in the oxidation states of technetium compounds. The Tc L₃-edge spectra are compared with those obtained from Mo and Ru oxide standards with various oxidation states and coordination environments. Most importantly, fitting the Tc L₃-edge to component peaks can provide direct evidence of crystal field splitting that cannot be obtained from the Tc K-edge.

1. Introduction

The 4d transition metal oxides continue to be one of the most exciting areas of solid-state chemistry, exhibiting a wide range of interesting magnetic and electronic properties. Of the 4d transition metal oxides, the structural and electronic properties of technetium oxides have remained somewhat elusive. This is not surprising given the absence of stable technetium isotopes and the radioactivity of the existing isotopes (Deutsch et al., 1983). One isotope in particular, ⁹⁹Tc, is a common byproduct in nuclear fission that decays by β-emission with a relatively long half-life of 2.12 × 10⁵ years. As such, there has been considerable effort to develop solid-state forms (e.g. pyrochlores) of technetium in order to sequester technetium from nuclear waste (Hartmann et al., 2011, 2012). Technetium (5s²4d⁵) is known to adopt a wide range of oxidation states analogous to manganese (4s²3d⁵) and rhenium (6s²5d⁵) (Kotegov et al., 1968). Technetium oxides are also known to exhibit magnetic properties that are unusual for 4d and 5d transition metal oxides (Rodriguez et al., 2011; Avdeev et al., 2011). Recently the perovskites CaTcO₃ and SrTcO₃, which adopt an orthorhombic structure in space group Pnma, were shown to exhibit antiferromagnetic ordering with exceptionally high Néel temperatures of 800 and 1023 K, respectively (Rodriguez et al., 2011; Avdeev et al., 2011). The magnetic moment on the Tc⁴⁺ cations in SrTcO₃ was relatively unaffected by a high-temperature orthorhombic cubic phase transition at 800 K (Thorogood et al., 2011a). It has been suggested that the unusual magnetic ordering in these materials originates from the half-filling of the t_2g band (Avdeev et al., 2011; Middey et al., 2012; Mravlje et al., 2012).

Because of the radioactive nature of technetium, there are relatively few tools available for analyzing the electronic properties of these materials compared with the other 4d transition metals. Tc K-edge X-ray absorption near-edge structure (XANES) has proven to be an effect tool for studying the oxidation state and coordination environment of Tc (Antonini et al., 1983; Almahamid et al., 1995; Mausolf et al., 2011; Poineau et al., 2008; Ferrier et al., 2012; Thorogood et al., 2011b). However, the Tc K-edge XANES spectrum, which predominantly corresponds to the dipole transition of 1s electrons into empty 5p states, is limited by a broad line shape and can only provide indirect information on the Tc 4d bonding states via the unoccupied p component in p–d hybridized orbitals. We have previously shown that the L₃-edge, which corresponds to the transition of 2p_3/2 electrons into unoccupied 4d states, is a sensitive probe for directly analyzing the bonding states of 4d transition metal oxide with empty or partially filled d orbitals (Zhou et al., 2009; Ting et al., 2010; Ricciardo et al., 2010; Qasim et al., 2011; Blanchard et al., 2012; Reynolds et al., 2013). In these studies, we demonstrated that crystal field splitting can be extracted from the 4d transition metal L₃-edge by fitting the spectra to component peaks (it is difficult to extract the crystal-field-splitting energy from the K-edge since the K-edge spectra do not directly probe the metal 4d states). To our knowledge, the Tc L-edge has never been measured. This is likely to be due to the radioactive nature of technetium (limiting the amount and form of technetium that can be analyzed at a synchrotron facility) and the inherent safety difficulties in conducting a Tc L-edge XANES experiment (the relatively weak excitation energy of the Tc L-edge at ∼2680 eV means that samples would have to be directly exposed to the incoming X-ray beam without any shielding).

Herein, we report on a method for measuring the Tc L₃-edge of technetium oxides that involves immobilizing powder samples in an epoxy resin. Although immobilizing samples in epoxy is a common practice in electron microscopy, it usually involves a much larger quantity than the allowed milligram quantity for Tc-containing samples. Furthermore, immobilizing powder samples in an epoxy resin is not common practice in soft X-ray experiments due to the relatively short penetration depth of soft X-rays compared with hard X-rays. It has previously been demonstrated that soft X-ray absorption spectroscopy experiments can be conducted on actinides (uranium- and thorium-based chlorides) by embedding the powder in a polystyrene matrix (Kozimor et al., 2009); in that work, the compounds of interest and the polystyrene were dissolved in a volatile solvent and the solid formed by evaporation of the solution. Since it is not possible to dissolve many condensed materials such as SrTcO₃, an alternative approach has been employed to study the Tc L₃-edge of SrTcO₃ (Tc⁴⁺) and NH₄TcO₄ (Tc⁷⁺) using an epoxy resin to immobilize the samples rather than polystyrene. We will demonstrate that both fluorescence (FLY) and total electron (TEY) yield measurements can be collected using this sample preparation method. These spectra will be compared with their respective Tc K-edge XANES spectra as well as the L₃-edge of other 4d transition metal oxides, specifically the neighboring Mo and Ru. The goals of this work are, firstly, to determine if spectra collected from epoxy-embedded samples are in fact representative of the bulk material and, secondly, to establish if information on the electronic and local structure of Tc cations can be obtained from the Tc L₃-edge XANES spectra.

2. Experimental

Caution! ⁹⁹Tc is a β⁻ emitter (E_max = 0.29 MeV). Appropriate shielding was employed during the synthesis and all manipulations. The purity of all starting materials was greater than 99%. NH₄TcO₄ was obtained from Oak Ridge National Laboratory and used as received. Although the sample was black in color, X-ray diffraction measurements indicated the sample to be phase pure. Polycrystalline samples of SrTcO₃ were synthesized using solid-state methods described elsewhere (Thorogood et al., 2011a). Stoichiometric amounts of NH₄TcO₄ and Sr(NO₃)₂ were dry rolled in polyethylene vials for 16 h, to ensure complete mixing, and annealed at 973 K under Ar for 1 h. The samples were re-ground, wet ball-milled in cyclohexane for 16 h, air-dried, pressed into pellets and annealed under Ar at 1423 K for 4 h. The purity of the samples was confirmed by powder X-ray diffraction (see supporting information¹).

All L₃-edge XANES spectra were collected on the soft X-ray beamline at the Australian Synchrotron (Cowie et al., 2010). In order to comply with the Australian Radiation Protection and Nuclear Safety Agency regulations, the amount of technetium used in these measurements was limited to ∼5 mg per sample. Technetium-containing powder samples were mixed with a minimum amount of industrial epoxy resin forming an extremely thick paste. The mixture was left to harden overnight by which time the technetium compounds were effectively immobilized. Swipe tests of the immobilized samples gave (beta) values of less than 100 counts s⁻¹ A similar method was used to prepare epoxy-embedded CaRuO₃ (Fig. 1). Ru and Mo oxide standards were lightly dusted onto double-sided carbon tape (SPI Supplies).

Figure 1 (a) Diagram of a powder sample embedded in an epoxy resin mounted onto a sample stage and (b) an image of the CaRuO3 sample used in the experiment. The diameter of the epoxy disc is approximately 1.1 mm.

All samples were inserted into the vacuum chamber via a load lock. The pressure inside the analysis chamber was maintained at <10⁻⁹ mbar during the measurements. The epoxy encapsulated samples did not significantly degas. Spectra were collected from ∼30 eV below to ∼80 eV above the edge, with a step size of 0.2 eV using both TEY and FLY modes (except that the FLY signal at the Ru L₃-edge had to be collected from another beamline, see §3). All spectra were recorded simultaneously with the TEY signal measured from Mo (Mo and Tc L₃-edge) or Ru(OH)Cl₃ (Ru L₃-edge) reference samples positioned upstream in the beamline. The reference sample removed approximately 10% of the beam intensity. Mo L₃-edge spectra were calibrated to the Mo L₃-edge of Mo foil, with the maximum of the first derivate set to 2520 eV. Tc L₃-edge spectra were calibrated to the Mo L₂-edge of Mo foil, with the maximum of the first derivate set to 2625 eV. Due to the relatively weak edge jump, the Ru L₃-edge spectra were calibrated against the maximum peak height of the Ru L₃-edge XANES spectra of Ru(OH)Cl₃ rather than the maximum of the first derivative. The maximum peak height of the Ru L₃-edge XANES spectra of Ru(OH)Cl₃ was at 2840.1 eV, based on the calibration against Ru metal with the maximum of the first derivate set to 2838 eV.

Tc K-edge XANES spectra of SrTcO₃ and NH₄TcO₄ were collected in transmission mode on the X-ray absorption spectroscopy beamline at the Australian Synchrotron (Glover et al., 2007). Powder samples were mixed with an appropriate amount of BN, sandwiched between two Kapton tapes and positioned directly in front of the X-ray beam. Spectra were collected from ∼200 eV below to ∼1000 eV above the edge with a step size of 0.25 eV around the edge and a dwell time of 1 s. X-rays were monochromated using an Si(111) monochromator, which was detuned by 50% to reject higher harmonics. The Tc K-edge was calibrated against Mo foil with the maximum of the first derivative of the Mo K-edge set to 20000 eV.

All absorption-edge energies are reported from the maximum of the first derivative of their respective edges. All XANES data were analyzed using the Athena and CasaXPS software packages (Ravel & Newville, 2005; Fairley, 2012).

3. Results and discussion

Measurement of the Tc L-edge spectra utilized the ultra-high-vacuum soft X-ray beamline at the Australian Synchrotron. The essential safety requirements for the measurements were (i) the sample was encapsulated in such a way that no loose material could be inadvertently introduced into the beamline and (ii) the amount of radioactive material was within the regulated limits of a general user beamline. For high-energy X-ray experiments, X-ray transparent windows can be employed to contain the sample; however, these are unsuit­able for low-energy soft X-ray measurements. Our strategy was to form a solid pellet of the technetium compound using an epoxy host. Before measuring the Tc L₃-edge, a proof-of-concept experiment was conducted to ensure that this sample preparation technique produces quality XANES spectra representative of the bulk material. For this, the Ru L₃-edge XANES spectrum of CaRuO₃ was collected. This was chosen because its absorption-edge energy is greater than that of the Tc L₃-edge and is near the upper limit of the energy range of the soft X-ray spectroscopy beamline (Cowie et al., 2010). Consequently, the X-ray signal at the Ru L₃-edge would be much weaker than that at the Tc L₃-edge.

The Ru L₃-edge XANES spectra of CaRuO₃, measured both as a powder dusted on carbon tape and as an epoxy-embedded sample, are shown in Fig. 2. Overall, the Ru L₃-edge spectra shown here are similar to those previously reported for CaRuO₃ (Zhou et al., 2009; Wu et al., 2003; Bréard et al., 2007). The Ru L₃-edge XANES spectrum of CaRuO₃, where Ru occupies the slightly distorted octahedral site, consists of two features that correspond to the transition of a 2p_3/2 electron into the unoccupied 4d–t_2g (lower energy peak) and 4d–e_g (higher energy peak) states. Such features are typically observed in 4d transition metal L-edge XANES spectra (Ting et al., 2010; Qasim et al., 2011; Blanchard et al., 2012; Reynolds et al., 2013; Blanchard et al., 2013a). The peak corresponding to transitions to the t_2g states appears as a lower energy shoulder due to the partial occupancy of the t_2g states in Ru⁴⁺ (t_2g⁴e_g⁰). As shown in Fig. 2, the Ru L₃-edge spectrum of the powder form of CaRuO₃ is nearly identical to that of the epoxy-embedded form. The crystal field splitting, estimated from fitting the Ru L₃-edge to component peaks corresponding to the t_2g and e_g peaks, was estimated to be ∼3.0 eV for both samples (more details later in the paper). Both spectra are similar to the FLY Ru L₃-edge spectrum of CaRuO₃ collected at the National Synchrotron Radiation Research Center in Hsinchu, Taiwan (see supporting information), confirming that the TEY signal is representative of the bulk material and that XANES spectra can be obtained using this sample preparation technique. We were unable to collect the FLY spectra on the soft X-ray spectroscopy beamline at the Australian Synchrotron because the Ru L₃-edge is outside the designed range of the FLY detector.

Figure 2 Ru L3-edge XANES spectra of powder (black line) and epoxy-embedded (red line) CaRuO3. The difference plot (green line) shows that the spectra are essentially identical. Both spectra were collected in TEY mode.

Our preliminary experiment with CaRuO₃ suggests that embedding powder samples in an epoxy resin is a feasible method for measuring the Tc L₃-edge. Consequently, we collected the Tc L₃-edge spectra from ∼5 mg SrTcO₃ and NH₄TcO₄ samples embedded in epoxy, and compared them with their respective K-edge XANES spectra (Fig. 3). Absorption-edge energies are tabulated in Table 1. Both Tc L₃-edges presented in Fig. 3 were collected in TEY mode. As shown in Fig. 4, the FLY signal is considerably weaker than the TEY signal. The FLY signal has a similar line shape to the TEY signal, confirming that the TEY signal of the Tc L₃-edge is representative of the bulk material and suitable for this analysis. The K and L edges have advantages and disadvantages when it comes to characterizing the electronic and structural properties of Tc oxides. Analysis of the Tc K-edge absorption-edge energy is particularly useful in identifying Tc⁴⁺ and Tc⁷⁺ species (Mausolf et al., 2011; Ferrier et al., 2012). The Tc L₃-edge absorption-edge energy of SrTcO₃ (2677.0 eV) is less than that of NH₄TcO₄ (2678.5 eV), but the difference between them (1.5 eV) is much less than the difference of 4.5 eV observed at the Tc K-edge (SrTcO₃, 21058.2 eV; NH₄TcO₄, 21062.7 eV). Both energy shifts are consistent with the Tc cations in SrTcO₃ having the lower oxidation state.

Table 1
Absorption-edge energies for the various 4d transition metal oxides

Sample	Oxidation state	Coordination number	Electronic configuration	Absorption edge	Absorption edge energy (eV)
(NH4)TcO4	Tc^7+	4	[Kr] 4d^0	Tc L3-edge	2678.5
				Tc K-edge	21062.7
SrTcO3	Tc^4+	6	[Kr] 4d^3	Tc L3-edge	2677.0
				Tc K-edge	21058.2
Bi2MoO6 (LT)	Mo^6+	6	[Kr] 4d^0	Mo L3-edge	2523.0
Bi2MoO6 (HT)	Mo^6+	4	[Kr] 4d^0	Mo L3-edge	2523.0
MoO2	Mo^4+	6	[Kr] 4d^2	Mo L3-edge	2521.4
Sr2RuYO6	Ru^5+	6	[Kr] 4d^3	Ru L3-edge	2839.1
SrRuO3	Ru^4+	6	[Kr] 4d^4	Ru L3-edge	2838.2
CaRuO3	Ru^4+	6	[Kr] 4d^4	Ru L3-edge	2838.1

Figure 3 (a) Tc L3-edge (collected in TEY mode) and (b) K-edge (collected in transmission mode) XANES spectra of SrTcO3 (octahedral Tc^4+) and NH4TcO4 (tetrahedral Tc^7+). Dashed lines correspond to the absorption-edge energies of SrTcO3 (black line) and NH4TcO4 (red line).

Figure 4 Tc L3-edge XANES spectra of SrTcO3 collected in TEY (black line) and FLY (red line) mode, respectively.

The most striking feature of the Tc L₃-edge is the line shape. The Tc L₃-edge line shape of NH₄TcO₄ consists of two features (labeled A and B in Fig. 3a). As previously discussed, features in 4d transition metal L-edge spectra typically result from the crystal field splitting of the 4d states. The Tc⁷⁺ cations have a tetrahedral coordination environment in NH₄TcO₄ and the features in the L₃-edge correspond to transitions to the empty e (feature A) and t₂ (feature B) states. Similar features corresponding to transitions to the partially occupied t_2g and e_g states are observed in SrTcO₃ with the e_g states undergoing further splitting due to distortion of the TcO₆ octahedra. This will be discussed further below. Evidence of crystal field splitting is not obvious in the Tc K-edge spectra, which is expected as the K-edge spectra do not probe the 4d states directly. Instead the Tc K-edge spectrum consists of a pre-edge feature which corresponds to the dipole forbidden transition of a 1s electron into 4d states via p–d hybridization (Kettle, 2008). In the Tc K-edge spectrum of NH₄TcO₄, the pre-edge contains an intense peak because the tetrahedral site occupied by the Tc⁷⁺ cation has T_d symmetry which allows the hybridization of metal 4d and 5p states (Kettle, 2008). In contrast, the Tc K-edge spectrum of SrTcO₃, where the Tc⁴⁺ cation occupies a nearly centrosymmetric octahedral site (O_h), does not contain an obvious pre-edge feature due to the lack of d–p hybridization (as there are no irreducible representations of O_h to which both p and d orbitals belong) (Kettle, 2008). A pre-edge feature at the K-edge can manifest in highly distorted octahedral systems because the reduced symmetry would allow a certain degree of p–d mixing (Yamamoto, 2008). Since the pre-edge feature results from a transition to the p component in d–p hybridized orbitals (not d-orbitals directly), it is difficult to extract information on crystal field splitting from 4d transition metal K-edge XANES spectra. The features at the Tc K-edge are also broader than those at the L-edge due to lifetime broadening. In general, the L-edge spectra appear to be more sensitive to changes in crystal field splitting, therefore having a greater sensitivity to changes in the local structure.

Since these appear to be the first measurements at the Tc L₃-edge, the experimental XANES spectra of SrTcO₃ (Tc⁴⁺) and NH₄TcO₄ (Tc⁷⁺) were compared with appropriate Mo and Ru L₃-edge standards (Fig. 5) measured on the same beamline. The difference in the Tc L₃-edge absorption edge energy between the Tc⁴⁺ and Tc⁷⁺ species (1.5 eV) is similar to that observed at the Mo L₃-edge between the Mo⁶⁺ and Mo⁴⁺ species (∼1.6 eV; Veith et al., 2005) and larger than that at the Ru L₃-edge between Ru⁴⁺ and Ru⁵⁺ oxides (0.9 eV). The energy shift in the Tc L₃-edge is somewhat smaller than expected for such a large change in oxidation state. Although there is no difference in the Mo L₃-edge absorption-edge energy of the high-temperature (tetrahedral Mo⁶⁺) and low-temperature (octahedral Mo⁶⁺) polymorphs of Bi₂MoO₆ (2523 eV), we cannot confirm from this study that the coordination environment of Tc has no effect on the Tc L₃-edge absorption-edge energy. Regardless, the Tc L₃-edge absorption-edge energy does appear to be sensitive to the oxidation state of Tc. The line shape of the Tc L₃-edge shares similarities to that of the Mo and Ru L₃-edge spectra. In general, features in XANES L₃-edge spectra of 4d⁰ transition metals correspond to the crystal field splitting of the 4d states. This is well illustrated by comparing the Mo L₃-edge spectra of the high-temperature and low-temperature polymorphs of Bi₂MoO₆ (Sankar et al., 1995; Buttrey et al., 1994; Laarif et al., 1984). The Mo L₃-edge spectra of both forms of Bi₂MoO₆ consist of two features (labeled A and B in Fig. 5a), which correspond to transitions to the doubly degenerate e and triply degenerate t₂ states in tetrahedral systems, and to the triply degenerate t_2g and doubly degenerate e_g states in octahedral systems, respectively (Bare et al., 1993; Hu et al., 1995; Lede et al., 2002). The difference in intensity of the two features reflects the reversal of the orbital energies and the appropriate degeneracy of these (3:2 in octahedral systems versus 2:3 in tetrahedral systems, under simplified crystal field theory). The Tc L₃-edge of NH₄TcO₄ has a similar line shape to the high-temperature form of Bi₂MoO₆, consistent with the presence of a tetrahedral 4d⁰ transition metal.

Figure 5 (a) Mo and (b) Ru L3-edge XANES spectra of various Mo and Ru oxide standards. All spectra were collected in TEY mode.

The line shape of the Tc L₃-edge spectrum of SrTcO₃ (Tc⁴⁺) is similar to that of other octahedral 4d transition metals with partially occupied 4d states, such as MoO₂ (Mo⁴⁺ d²) and SrRuO₃ (Ru⁴⁺ d⁴). As mentioned earlier, the intensity of feature A is weaker than that of feature B due to partial filling of the t_2g state. The splitting between the t_2g and e_g orbitals depends on the ligand field strength, which typically increases as the oxidation state increases (with similar bond distances). The transition into the t_2g states appears as a discrete peak (labeled A in Fig. 5b) in the L₃-edge spectra of Sr₂Ru^VYO₆ but only as a low-energy shoulder in Sr(Ca)Ru^IVO₃ and SrTc^IVO₃, consistent with the higher Ru oxidation state in Sr₂RuYO₆. In octahedral systems, feature B, which corresponds to the e_g states, shows asymmetry towards higher energy. This asymmetry has been observed in the Zr L₃-edge of pyrochlores Ln₂Zr₂O₇ (Blanchard et al., 2012), and is believed to be a consequence of the sensitivity of the e_g states to local distortions due to the overlap of d_z² and d_x²-y² states with ligand states. Such distortions can result in broadening and splitting of the e_g peak (Schneller et al., 2008).

Further information on the local structure can be obtained by fitting the Tc L₃-edge to component peaks. Such analysis has previously been employed to study the local structure of zirconate and hafnate pyrochlores (Qasim et al., 2011; Blanchard et al., 2012, 2013b; Reynolds et al., 2013). As shown in Fig. 6, the spectra were fitted to pseudo-Voigt peaks with an arctan background. Fitted spectra of the Ru and Mo L₃-edge XANES spectra can be found in the supporting information (Fig. S2 of the supporting information). Results of the fitting are presented in Table 2. In all compounds listed, feature A was fitted to a single component peak. The number of components required to fit feature B varied depending on the coordination environment. In compounds where the metal cation occupies a tetrahedral coordination environment, only one component peak was required to fit feature B. In compounds where the metal is six-coordinate, two component peaks (labeled B1 and B2) were required to fit feature B. This is likely to be due to localized distortions of the metal octahedra (Ray et al., 2011; Ikeno et al., 2013). The splitting of feature B into two peaks can be confirmed by analysing the first derivative XANES spectrum. For example, the first derivative Tc L₃-edge XANES spectrum of SrTcO₃ (see supporting information) has three peak maxima which corresponds to three unique peaks. By comparison, the first derivative Tc L₃-edge XANES spectrum of NH₄TcO₇ has only two peak maxima. In order to reduce the number of free parameters, peak widths were constrained so that all peaks in a spectrum had the same width. Crystal-field-splitting energy (ΔE) can be taken as the energy difference between peaks A and B. For octahedral systems, ΔE is taken as the energy difference between E_A (energy of peak A) and the weighted average of E_B1 and E_B2 (energies of peaks B1 and B2, respectively), i.e. E_B = (E_B1I_B1 + E_B2I_B2)/(I_B1 + I_B2). The ΔE value of NH₄TcO₄ (2.5 eV) is less than that of SrTcO₃ (2.9 eV). This is to be expected because tetrahedral systems have less crystal-field-splitting ΔE (in the simplified crystal field theory Δ_Tet ∼ 4/9Δ_Oct assuming the same oxidation state and distances). This reduction is evident when comparing the spectra of the low (oct) and high (tet) temperature forms of Bi₂MoO₆ (3.1 versus 2.0 eV). The ΔE value also scales with the oxidation state of the cation. For example, ΔE for SrRu^IVO₃ (2.8 eV) is less than that of Sr₂Ru^VYO₆ (3.1 eV). In the present technetium oxides these two opposing factors result in a larger splitting in the octahedral Tc⁴⁺ oxide compared (2.9 eV) with the tetrahedral Tc⁷⁺ oxide (2.6 eV).

Table 2
Fitting parameters to the 4d transition metal L₃-edges

E_i is the energy of peak i and I_i is the intensity of peak i estimated from peak fitting. The width of the peaks and the estimated splitting between peaks A and B1 (ΔE) are also given.

Sample	EA (eV)	EB1 (eV)	EB2 (eV)	IA	IB1	IB2	Peak width	ΔE (eV)
(NH4)TcO4	2679.5	2682.0	N/A	2.682	3.201	N/A	2.7	2.5
SrTcO3	2678.0	2680.0	2682.3	4.265	9.283	6.136	2.7	2.9
Bi2MoO6 (LT)	2524.0	2526.7	2529.0	10.534	9.37	2.19	2.5	3.1
Bi2MoO6 (HT)	2523.6	2525.9	N/A	3.766	11.377	N/A	2.4	2.0
MoO2	2523.0	2525.0	2527.4	4.887	8.168	5.046	3.3	3.2
Sr2RuYO6	2840.2	2843.1	2846.2	33.213	45.249	3.278	2.8	3.1
SrRuO3	2839.2	2841.2	2843.9	20.964	29.95	12.38	2.9	2.8
CaRuO3	2839.6	2841.9	2845.1	27.436	38.139	8.186	3.4	3.1

Figure 6 Fitted Tc L3-edge XANES spectra of (a) NH4TcO4 and (b) SrTcO3.

4. Conclusion

In conclusion, we have successfully demonstrated that the Tc L₃-edge can be measured by embedding a very small amount of technetium-containing powder samples (∼5 mg) in an epoxy resin. We have shown that the L-edge XANES spectra of the early 4d transition metals (Zr, Nb, Mo, Tc and Ru) are extremely sensitive to changes in both the oxidation state and coordination environment, providing information not only on the formal oxidation state of the cation but also on the crystal field splitting and coordination environment. Since the 4d orbitals are not fully occupied, the Tc L₃-edge spectrum directly probes the 4d bonding states, which can provide more accurate information on the local structure of Tc cations than the Tc K-edge spectrum. Crystal-field-splitting energies can be deduced from the Tc L₃-edge XANES spectra by fitting the spectra with component peaks. The methodology used here can easily be extended to other technetium species of various oxidation states and coordination environments, an important step in exploring this relatively unknown area of X-ray absorption spectroscopy.

Footnotes

¹Supporting information for this paper is available from the IUCr electronic archives (Reference: HF5261 ).