Three-component spectroelectrochemical sensor module for the detection of pertechnetate (TcO₄^-)

Introduction

One of the basic challenges for environmental remediation of radiological wastes is to monitor radiochemical constituents in various areas, ranging from the containment of low/medium- and high-level radioactive waste to contaminant plumes in subsurface water. Present methods of analysis are hazardous, expensive and time-consuming, usually requiring lengthy sampling, preparation, analysis and data interpretation. An alternative approach is to use sensors for the rapid, sensitive and economic in situ analyses for various constituents of interest (Monk et al. 2005).

Technetium (Tc) is one such constituent of radioactive waste where the need for a method of detection exists, but an efficient sensor does not. The element Tc is not found in appreciable quantities in nature. However, since the isotope ⁹⁹Tc is a byproduct of thermal nuclear fission of ²³⁵U, ²³³U, and ²³⁹Pu (at yields of 6.1%, 4.8%, and 5.9%, respectively) (Siegel 1946), significant quantities of ⁹⁹Tc are found at nuclear sites. The production of ⁹⁹Tc accounts for ∼100% of all isotopic sources of Tc. As a representative example, the total ⁹⁹Tc content in the US DOE site at Hanford, WA is ∼2000 kg, of which an estimated 4% has been lost to the environment (Figure 1) (Hartman et al. 2006). Therefore, one of the basic necessities in Tc remediation is to meet regulatory requirements for near term disposal of Tc found in DOE waste streams and released to the environment. Design of novel Tc separation and immobilization technologies requires detailed knowledge of Tc chemistry, including speciation, partitioning, retention, and transport in DOE flow sheets or within groundwater and soils.

Figure 1

Regional plume map showing ⁹⁹Tc in the vicinity of the Hanford Site. The large plume marks the Tc found in groundwater (units in pCi/l). The plume is migrating northwest and has reached the Columbia River and has a source in the 200-East Area of the Hanford Site (Hartman et al. 2006).

While ⁹⁹Tc exhibits a rather weak β^- decay (E_max= 0.292 keV), it is of a high concern primarily for two reasons: (a) the long half-life (2.13×10⁵ years) and (b) the high solubility of its most abundantly occurring natural form, pertechnetate (TcO₄^-), in aqueous environmental media (∼126 g l^-1) (Schwochau 2000). Also, due to its negative charge, TcO₄^- has very low retention in the soil and migrates with groundwater (Strickert et al. 1980, Elwear et al. 1992, Schulte and Scoppa 1987).

TcO₄^- does not have a readily identifiable optical fingerprint in the low-energy visible region, making its detection and identification highly challenging in the presence of other potential interfering constituents that mostly absorb light in the higher-energy visible region. For example, the potential interfering species in the nuclear wastes in Hanford, WA consist of anionic interferents (nitrate, chloride, sulfate, ferrocyanide and chromate) as well as other constituents (uranium, ⁶⁰Co, sodium and calcium); some of these have strong absorption in the same spectral region where TcO₄^- absorbs, making identification and detection of TcO₄^- highly difficult. Furthermore, since ⁹⁹Tc decays through release of low energy β^-, it cannot be monitored remotely through steel lined boreholes. The current methods of TcO₄^- analysis, including ion chromatography (IC), mass spectrometry (MS), IC-MS, capillary electrophoresis, ion selective electrodes, Surface Enhance Raman Spectroscopy (SERS), Extended X-ray Absorption Fine Structure (EXAFS) spectroscopy or combinations of these techniques, suffer from limitations of often dangerous sample collection, imperfect selectivity, and long processing times. To overcome these challenges, we focus on designing an in situ spectroelectrochemical sensor that combines several components of detection. This review describes the advances of the spectroelectrochemical detection of Tc.

Three component sensor

For the last several years, our research group has been developing a spectroelectrochemical sensor that combines the functional components of (a) selective partitioning, (b) electrochemistry and (c) optical spectroscopy into a single device (Shi et al. 1997a,b, Gao et al. 1999, Gao et al. 2001, Slaterbeck et al. 2000, Slaterbeck et al. 1999, DiVirgilio-Thomas et al. 2000, Maizels et al. 2000, Maizels et al. 2002, Ross et al. 2000, Richardson et al. 2002, Heineman et al. 2003, Kaval et al. 2003, Stegemiller et al. 2003). This sensor is based on a novel device that incorporates multiple internal reflection (MIR) spectroscopy in an optically transparent electrode (OTE) coated with a selective film (Kaval et al. 2003). The film is a thin, solid polymeric polyelectrolyte and is an essential element in the operation of the sensor. It is coated onto the surface of the OTE sensing platform supported on a glass slide and has dual functionalities; (a) it facilitates chemical- and charge-selective transport of ions to the electrode and (b) it enhances the detection limit by preconcentrating the analyte into its matrix. The concept is illustrated by the diagram shown in Figure 2. When light propagates through the glass waveguide, it undergoes multiple internal reflections. At each reflection point, an evanescent field is generated which penetrates the film, and as a result, the reduced or oxidized form of the analyte can be monitored optically. In its operation, an electrical potential is applied to the OTE to cause electrolysis of selective analytes that have partitioned into the film; this helps to enhance selectivity, as only those analytes that are redox active at the applied potential get oxidized or reduced. Quantification of the analyte is based on the magnitude of the change in the optical signal, which is proportional to the concentration of analyte in the film as defined by the partition coefficient and, in turn, is proportional to its concentration in the liquid sample. The film can also be used as an embedding matrix for carefully selected ligands that have the ability to react readily with Tc, to form coordination complexes with highly pronounced spectroscopic signatures that are more electroactive within the range of the studied potentials. The coordination can proceed by either (a) electrochemical reduction of Tc(VII) followed by the binding of the ligands or (b) through chemical reduction of Tc(VII) by substrates embedded in the polymer matrix. The use of the ligands in the film serves the following advantages: (a) the ligands can coordinate with the Tc to form complexes in situ that have more prominent/easily identifiable spectroscopic signatures in the visible region, (b) the formation of the coordination complexes can generate a more sensitive detection mode having a lower limit of detection {as a representative example, while TcO₄^- does not exhibit any luminescence within 400–800 nm, [Tc(dmpe)₃]²⁺ [dmpe=1,2-bis-(dimethylphosphino)ethane] exhibits a strong red luminescence at ∼660 nm and therefore utilizes a detection mode that is unavailable to TcO₄^-}, (c) choosing ligands which exhibit a stronger preferential binding affinity towards Tc compared to other metal centers can help to screen out a significant number of interfering metal centers, and (d) certain ligands can also generate an electroactive species and further help in the selection process.

Figure 2

Schematic representation of a three component sensor and its operation. The polymer screens out analytes of dissimilar charge, the ITO layer discriminates among analytes based on their redox potential, and the transmitted light activates analytes based on their absorption/emission.

This sensor allows three modes of selectivity and is therefore particularly useful in detection of analytes in the presence of various interfering species.

• The first parameter of selectivity is enforced by the ion-exchange polymer-polyelectrolyte film, which facilitates the partitioning of either negatively or positively charged species. By selecting the charge on the polymer backbone, one can selectively chose partitioning of ions of a particular charge, while those of the opposite charge are screened out.

• The second parameter of selectivity is introduced by the potential applied to the OTE. One can carefully apply a potential at which only the analytes of interest are electroactive and are selectively electrolyzed, while the others remain unchanged.

• The third parameter of selectivity is based on optical spectroscopy (absorption and/or emission). Light under multiple internal reflections at an interface establishes an evanescent field which penetrates about one wavelength into the film (Figure 3) (Fornel 2001, Kaval et al. 2003). If an analyte partitions into the film, and if it absorbs or luminesces when excited by the light propagated in the waveguide, the transmitted light is attenuated and gives a signal response. Thus, only the analyte(s) that result(s) in attenuation of the signal will be selectively measured.

Figure 3

Principle of evanescent wave. Adapted with permission from “Kaval, N.; Seliskar, C.J.; Heineman, W.R. Anal. Chem. 2003, 75, 6334” Copyright © 2003, American Chemical Society.

The three modes of selectivity are particularly attractive for the specific detection of analytes in presence of multiple interfering species, where a one- or two-parameter selectivity is unable to selectively identify and detect a single analyte of interest.

In addition to selectivity, an important consideration for the effective performance of the sensor is the sensitivity. It has been seen that detection mode plays an important role in the sensitivity. In the studies that have used absorption spectroscopy (“sensor absorbance”) (Shi et al. 1997a,b, 1998, Slaterbeck et al. 1997, 1999, 2000, Gao et al. 1999, 2001, DiVirgilio-Thomas et al. 2000, Maizels et al. 2000, 2002, Richardson et al. 2002, Heineman et al. 2003, Kaval et al. 2003, Stegemiller et al. 2003) as the mode of optical detection, sensitivity was determined in part by the two parameters, the effective path length (Mendes et al. 2000, Fornel 2001) and the change in molar extinction coefficient (Δε) that accompanies electrolysis of the light-absorbing species. In the case of multiple internal reflections, the effective path-length can be determined (Fornel 2001) by the penetration depth of the evanescent field into the selective film and the number of reflections that the propagating light undergoes. A third parameter affecting sensitivity is the partition coefficient of the analyte from the solution into the selective film, which determines the concentration enhancement in the optical path length that the film provides. Since electrochemistry is used only for modulation of the optical signal, the usual Faraday’s law parameters are not a factor in determining sensitivity. Improvements in the sensitivity generally translate into a lower limit of detection. Considering these factors, the limits of detection in the 10^-6–10^-5 m range have been observed for the analytes with ε values of ∼1000 l mol^-1 cm^-1 (e.g., [Fe(CN)₆]^3-) using sensors based on absorbance mode of detection) (Maizels et al. 2002). By contrast, in the species exhibiting emission properties, the limit of detection can be significantly lowered. Therefore, redox active luminescent probes such as [Ru(bpy)₃]²⁺ (where bpy=2,2′-bipyridine) and [Re(dmpe)₃]²⁺ allow for the direct detection of the analyte as low as ∼10^-10–10^-13 m (Chatterjee et al. 2011b, Kaval et al. 2003). With improvements in the collection and excitation optics of the cell, even lower detection limits are possible. More importantly, the sensing volume can be sufficiently low. As a representative example, volumes of the order of ∼4 μl have proved to be sufficient for efficient and reliable detection (Kaval et al. 2003). Moreover, using emission as a detection mode can help discriminate the non-luminescent interfering species present in the analyte solution. These factors highlight the advantages of choosing emission as the detection mode, and emphasize the importance of choosing the proper set of ligands to coordinate with the Tc metal center.

Choice and development of polymer films

The polymer films constitute are an important constituent of the sensor performance and play a significant role in controlling both the selectivity and the sensitivity of the sensor. The sensitivity is primarily based on the preconcentration ability of the film, which is related to the physicochemical properties of the polymer films (Deng et al. 1997, Chatterjee et al. 2011b). The charge selectivity becomes important when the uptake of ions into the polymer is governed by an ion exchange mechanism rather than ion-pair formation. The chemical selectivity can be related to the lipophilicity of the polymer and the polarizability of the functional groups.

Films for entrapment of cationic species. Partially sulfonated polymers (e.g., Nafion, polystyrene sulfonate (PSS) and polystyrene-block-poly(ethylene-ran-butylene)-block-polystyrene (SSEBS), shown in Figure 4, take up bulky positively charged cations. Both SSEBS and Nafion have demonstrated the ability to preconcentrate a range of cations, both organic (e.g., phenosafranine and rhodamine 6G) and inorganic coordination compounds (e.g., [Ru(bpy)₃]²⁺, [Fe(bpy)₃]²⁺, [Cu(bpy)₂]²⁺) (Ross et al. 2000, Andria et al. 2004, 2009, Pantelic and Seliskar 2007; Pantelic et al. 2006, 2009). More importantly, the spectroscopic properties (luminescence intensity and emission life-times) of these compounds are not destroyed or significantly reduced by the films upon their entrapment and can be exploited for their detection and modulation. This property is particularly attractive for compounds that retain their luminescence properties even after their entrapment into the films, as the emission intensity can be effectively used as the detection modality instead of absorbance, which can significantly lower the detection limit. This was true for [Ru(bpy)₃]²⁺ where using the phosphorescence from the molecule as the detection modality improved the detection limit by five to six orders of magnitude (DL=5×10^-12–10^-13 m for emission, 5×10^-7 m for absorbance).

Figure 4

Representative cation exchange polymer films.

The SSEBS films also exhibit the charge selectivity that is essential for chemical sensing ability of sensors fabricated with this novel thin film material. As an example, demonstration of the selective partitioning mode using binary mixtures of model analytes like a [Ru(bpy)₃]²⁺/[Fe(CN)₆]^3- mixture, shows that the negatively charged [Fe(CN)₆]^3- is excluded from the SSEBS film, and therefore does not interfere with the response of the sensor to the [Ru(bpy)₃]²⁺. This is suggestive that the uptake of migrating ions by these films occurs mechanistically through anion exchange, rather than ion-pair formation. Furthermore, the importance of using three modes together can be demonstrated by the analysis of [Ru(bpy)₃]²⁺/[Fe(bpy)₃]²⁺ and [Ru(bpy)₃]²⁺/[Cu(bpy)₂]²⁺ test mixtures, where both of the analytes can penetrate into the film. Here, both selection of a specific wavelength for absorption and selection of a specific potential window are required to separate out or eliminate the signal of the analyte to be detected from the interfering analyte. Finally, analysis of the [Ru(bpy)₃]²⁺/[Fe(bpy)₃]²⁺ test mixture was also demonstrated using luminescence detection (Andria et al. 2009). This technique for efficient detection of multiple analytes can be applied even to systems where a more involved mode of detection has to be employed. As a representative example, the SSEBS film was used to study the ability of a spectroelectrochemical sensor to simultaneously detect two analytes using model analytes Pb²⁺ and Fe²⁺; in the system, Pb²⁺ was detected by optical stripping voltammetry and Fe²⁺ was detected in situ by ligand complexation followed by redox modulation of optical absorbance (Andria et al. 2010).

Films for entrapment of anionic species. The proper selection of films in the sensor for the entrapment of anions depends not only on their ability to allow TcO₄^- diffusion and preconcentration into the film, but also on the voltammetric behavior of the preconcentrated TcO₄^- entrapped inside these films. Poly(vinylbenzyl trimethylammonium chloride) (PVTAC), poly(dimethyldiallylammonium chloride) (PDMDAAC, where polymerization results in internal cyclization to form a five membered ring, as shown in Figure 5), or quaternized poly(4-vinylpyridine) (QPVP) (shown in Figure 5), or their hybrids PVTAC-PVA (PVA=polyvinyl alcohol), PDMDAAC-SiO₂ or QPVP-SiO₂, have been particularly promising in sensor applications.

Figure 5

Representative anion exchange polymer films.

Choice of the optically transparent electrode

Voltammetry of TcO₄^- at OTE surfaces. In the sensor development, careful consideration has to be provided to the choice of the OTE in terms of it being both (a) optically transparent and (b) redox inactive in the window studied. Using Indium Tin Oxide (ITO) as the optically transparent electrode has produced promising results for TcO₄^- systems (Monk et al. 2005). The electrochemistry of TcO₄^- on ITO surfaces has shown that the voltammetric behavior of TcO₄^- on ITO is very similar to its behavior on a platinum working electrode (Mazzocchin et al. 1974), which is the conventionally accepted working electrode of choice for TcO₄^- electrochemistry. In addition, ITO has a large enough spectroscopic and electrochemical window allowing for the efficient monitoring of the necessary spectroscopic and redox changes in TcO₄^-.

Voltammetry at bare ITO surface. Voltammograms of aqueous TcO₄^- solutions at bare ITO show poorly defined reduction waves for TcO₄^- (shown in Figure 6). The reduction of TcO₄^- occurs between -0.80 V and -1.00 V as a cathodic shoulder on a larger wave for the evolution of H₂ gas from water. The current increases with the evolution of hydrogen, and on the return scan, a crossing over the initial scan occurs around -0.90 V, indicative of a changing electrode surface due to the electrodeposition of TcO₂.

Figure 6

Cyclic voltammograms of pertechnetate (TcO₄^-) in pH 7 phosphate buffer as supporting electrolyte, 25 mVs^-1 scan rate, Ag/AgCl reference electrode, Pt auxiliary electrode, solution deoxygenated for 30 min prior to scan: (A) 5.0×10^-4 m TcO₄^- on bare ITO electrode; and (B) 1.6×10^-3 m TcO₄^- on ITO electrodes coated with quaternized poly(4-vinylpyridine) (QPVP)-SiO₂, poly(vinylbenzyl trimethylammonium chloride)-polyvinyl alcohol (PVTAC-PVA) and poly(dimethyldiallylammonium chloride) (PDMDAAC)-SiO₂ after 30 min of immersion in sample. Reprinted with permission from “Monk, D.J.; Stegemiller, M.L.; Conklin, S.; Paddock, J.R.; Heineman, W.R.; Seliskar, C.J.; Ridgway, T.H.; Bryan, S.A.; Hubler, T.L.; “Spectroelectrochemical Sensor for Technetium: Preconcentration and Quantification of Pertechnetate in Polymer-Modified Electrodes,”. ACS Symposium Series 904 Subsurface Contamination Remediation: Accomplishments of the Environmental Management Science Program, (2005), 306–321; American Chemical Society: Washington, DC, 2005; Vol. 904” Copyright © 2005, American Chemical Society.

Reduction:

During the reverse scan, an anodic oxidation is observed at 0.15 V which is assigned to the oxidation of TcO₂ to TcO₄^- based on the similarity of the voltammogram to that obtained on a platinum working electrode surface (Mazzocchin et al. 1974).

Oxidation:

While the redox system is not chemically reversible (TcO₄^- is reduced to TcO₂ which eventually coats the bare electrode surface), the fact that TcO₄^- is electroactive at the ITO surface, and can undergo redox changes within the ITO potential window is critical for the selection of ITO as a component in the sensor.

Voltammetry at polymer modified ITO surface. Voltammograms of aqueous TcO₄^- solutions on ITO surfaces coated with PVTACS-PVA, PDMDAAC-SiO₂ or QPVP-SiO₂ polymers exhibit a TcO₄^- reduction which is more sharply defined compared to that produced on a bare ITO surface (Monk et al. 2005). There is also a significant increase in the currents for the redox processes. Furthermore, while bare ITO shows minimal TcO₂ deposited on the surface, there is a significant increase with an anion exchange film, and the amount of Tc deposited is proportional to the concentration of the solution, as observed in Figure 7. This figure also shows the correlation plot of increased optical density vs. solute concentration of TcO₄^-, indicating a direct proportionality of reduced Tc in the polymer to solution concentration of TcO₄^-. This is suggestive of effective preconcentration of TcO₄^- into the films, which is a critical requirement for lowering the limit of detection. The presence of the film also results in a mechanistic change in the TcO₄^- reduction, with the voltammograms appearing more chemically reversible. This is consistent with the formation of a relatively stable lower oxidation state of Tc instead of the insoluble TcO₂. The lower oxidation state is apparently more stabilized by the presence of the polymer film, possibly by electrostatic interactions with the positively charged quaternary ammonium groups through ion pairing, or through coordination of tertiary amine/imine nitrogens that remain coordinatively unsaturated. It is important to note that while there is a general enhancement of the TcO₄^- reduction in all three films, due to the electrostatic affinity for TcO₄^- anion of the quaternary ammonium ions, the voltammograms are different enough to indicate the importance of film choice.

Figure 7

(A) Photographs showing the significant increase of Tc deposition with an anion exchange film compared to a bare ITO surface: (1) 1.6×10^-3 m pertechnetate (TcO₄^-) on bare ITO; (2) 1×10^-4 m TcO₄^- on quaternized poly(4-vinylpyridine) (QPVP)-SiO₂; (3) 5×10^-4 m TcO₄^- on QPVP-SiO₂; and (4) 1.6×10^-3 m TcO₄^- on QPVP-SiO₂; and (B) plot of absorbance of TcO₂ vs. ⁹⁹Tc dose rate for three concentrations of TcO₄^- on QPVP-SiO₂. Reprinted with permission from “Monk, D.J.; Stegemiller, M.L.; Conklin, S.; Paddock, J.R.; Heineman, W.R.; Seliskar, C.J.; Ridgway, T.H.; Bryan, S.A.; Hubler, T.L.; “Spectroelectrochemical Sensor for Technetium: Preconcentration and Quantification of Pertechnetate in Polymer-Modified Electrodes,”. ACS Symposium Series 904 Subsurface Contamination Remediation: Accomplishments of the Environmental Management Science Program, (2005), 306–321; American Chemical Society: Washington, DC, 2005; Vol. 904” Copyright © 2005, American Chemical Society.

Development of sensor instrumentation

An ideal sensor device to be used for real-time in-field detection of radiological waste forms needs to be portable, robust and easy to operate with protective gear. In the basic configurations, the instrument consists of a spectroelectrochemical cell, a spectrophotometer, a light source and a portable computer (Chatterjee et al. 2011b). The components are shown in Figure 8. The cell is small enough (15.24×10.16×5.08 cm) to be easily transferable and accommodates optical, electrochemical and liquid sample introduction components. A sensor chip, a 2.54×2.54 cm ITO slide coated with a selective film, can be easily loaded since in the design, due consideration was made for its rapid assembling and disassembling by an experimenter wearing radioactivity protection equipment. Optical signals to and from the cell are carried by fiber optic cables. Liquid sample introduction to the cell can be done through thin Teflon tubing by manual injection or by using automated injection equipment. A three-wire cable is connected to the cell from a potentiostat for electrochemical modulation of the optical signal.

Figure 8

(A) Side view of the spectroelectrochemistry sensor; (B) overhead view of the optics module and the electrochemistry module; and (C) magnified side view of the electrochemistry module. Reprinted with permission from “Chatterjee, S.; Del Negro, A.S.; Edwards, M.K.; Bryan, S.A.; Kaval, N.; Pantelic, N.; Morris, L.K.; Heineman, W.R.; Seliskar, C.J. Anal. Chem. 2011, 83, 1766.” Copyright © 2011, American Chemical Society.

Detection of TcO₄^- and its surrogate perrhenate

The basic mode of detection of TcO₄^- anion and its surrogate ReO₄^- anion is based on preconcentration of the anion by a polymer film, followed by the reduction of the anion (chemically or electrochemically) to a lower oxidation state. Strongly coordinating ligands are also entrapped within the polymer matrix either prior to, or following the preconcentration of the TcO₄^- /ReO₄^- anion, which can coordinate with the Tc/Re metal center in situ once it is reduced to form a coordination complex. The ligands are chosen in such a way to ensure that the generated complex has a strong, easily identifiable and easily distinguishable spectral signature.

Tc-dioxo complexes

The trans-dioxo Tc(V) tetraimine complexes (two representative complexes shown in Figure 9) and their Re(V) dioxo surrogates have been known for several years now. Both the Tc(V) complexes and their Re(V) analogs can be generated easily from their respective TcO₄^-/ReO₄^- anions through straightforward non-elaborate two-step synthesis with a variety of N-donor ligands. A schematic for the synthesis of trans-dioxo Tc(V) tetraimine complexes from TcO₄^- is shown in scheme 1 (Kastner et al. 1984, Del Negro et al. 2005, Chatterjee et al. 2011a).

Figure 9

Representative trans-dioxo Tc(V) tetraimine complexes. Reprinted with permission from “Chatterjee, S.; Del Negro, A.S.; Wang, Z.M.; Edwards, M.K.; Skomurski, F.N.; Hightower, S.E.; Krause, J.A.; Twamley, B.; Sullivan, B.P.; Reber, C.; Heineman, W.R.; Seliskar, C.J.; Bryan, S.A. Inorg. Chem. 2011, 50, 5815.” Copyright © 2011, American Chemical Society.

Scheme 1

Synthetic scheme. Conditions: (i) Bu₄NCl, HCl; (ii) (a) pyridine, (b) 4-picoline. Reprinted with permission from “ Chatterjee, S.; Del Negro, A.S.; Wang, Z.M.; Edwards, M.K.; Skomurski, F.N.; Hightower, S.E.; Krause, J.A.; Twamley, B.; Sullivan, B.P.; Reber, C.; Heineman, W.R.; Seliskar, C.J.; Bryan, S.A. Inorg. Chem. 2011, 50, 5815.” Copyright © 2011, American Chemical Society.

The trans-dioxo Re(V)-tetraimine complexes have been well studied and show intense visible emission [as an example, trans-dioxo Re(V)-tetrapyridyl complexes show visible emission at ∼650 nm (Φ=2–4% in tetrahydrofuran)] (Winkler and Gray 1985). This emission is retained inside polymer matrices, which is a critical requirement for the sensor performance. trans-dioxo Tc(V) tetrapyridyl ([TcO₂(py)₄]⁺) and trans-dioxo Tc(V) tetrapicolyl ([TcO₂(pic)₄]⁺) complexes exhibit emissions arising from electronic configurations that are similar to the Re(V) complexes (Del Negro et al. 2005, Chatterjee et al. 2011a). Upon 415 nm excitation at room temperature, the complexes exhibit broad, structureless luminescence with emission maxima at approximately 710 nm {for [TcO₂(py)₄]⁺} and 750 nm {for [TcO₂(pic)₄]⁺}. Similar to the Re(V) analogs, the emission in the Tc(V) analogs arise from a Eg excited state (Hartman et al. 2006). The similarity in the luminescence properties of the Tc(V) complexes to the Re(V) analogs allows us to use the latter as surrogates for the Tc(V) complexes to optimize the sensor behavior. The emissions are red-shifted compared to the Re(V) analogs by 60–100 nm, which moves the emissions completely beyond the realm of autofluorescence of the polymer films, and the possible interfering species mentioned earlier. Upon cooling the complexes to 8 K, distinct vibronic features appear in the spectra, along with an increase in emission intensities, as shown in Figure 10. The low temperature emission spectra display the characteristic progressions of the symmetric O=Tc=O and Tc-L stretching modes. The red-shift of the Tc(V) emission compared to Re(V) is maintained at low temperatures.

Figure 10

Low temperature (8 K) luminescence spectra of microcrystalline: (A) [ReO₂(py)₄]Cl·2H₂O (

) and [TcO₂(py)₄]Cl·2H₂O (

); and (B) [ReO₂(pic)₄](BPh₄) (

) and [TcO₂(pic)₄](BPh₄) (

). Reprinted with permission from “Del Negro, A.S.; Wang, Z.M.; Seliskar, C.J.; Heineman, W.R.; Sullivan, B.P.; Hightower, S.E.; Hubler, T.L.; Bryan, S.A. J. Am. Chem. Soc. 2005, 127, 14978.” Copyright © 2005, American Chemical Society.

Tc(II) tris-phosphine complexes and congeners

In our search for new and better luminescent Tc complexes that can be made from TcO₄^-, we examined [Tc(dmpe)₃]⁺, which have previously been studied for nuclear medicine procedures. The congener [Tc(dmpe)₃]⁺, or other similar congeners, can be considered attractive targets, as these molecules are anticipated end products of the complete electrochemical reduction of TcO₄^- in the polymer film in the presence of phosphine ligands. Previous studies have shown that non-aqueous [Tc(dmpe)₃]⁺ is part of a redox couple with well-defined, reversible electrochemistry. It can be reversibly converted to the higher oxidation state [Tc(dmpe)₃]²⁺ according to the process below, with a formal reduction potential of 0.05 V vs. Ag/AgCl: (Doyle et al. 1986, Chatterjee et al. 2011b, 2012)

The process is accompanied by distinct absorption changes in the visible wavelength region. In aqueous solution, [Tc(dmpe)₃]⁺ absorbs at 250 nm but has no absorption in the visible region; [Tc(dmpe)₃]²⁺ also absorbs at 250 nm and has visible absorption at 585 nm (ε=1850 m^-1 cm^-1) (Lee and Kirchhoff 1994). The luminescence properties of [Tc(dmpe)₃]^2+/+ in aqueous solution and in films have also been thoroughly studied and have shown that while [Tc(dmpe)₃]⁺ is non-emissive, the higher oxidation state [Tc(dmpe)₃]²⁺ (generated by chemical or electrochemical oxidation) exhibits a strong emission at 660 nm (λ_ex=532 nm) (shown in Figure 11) which can be used as a detection modality (Del Negro et al. 2006, Chatterjee et al. 2011b, 2012). The molar extinction coefficient (ε=1850 m^-1 cm^-1; at 585 nm, acetonitrile) (Del Negro et al. 2006) of [Tc(dmpe)₃]²⁺ is very similar to that of [Re(dmpe)₃]²⁺ (ε=2110 m^-1 cm^-1 at 530 nm, acetonitrile) (Lee and Kirchhoff 1994) at its longest wavelength absorption, which is assigned to a σ(P)→dπ(Re) transition. The corresponding Stokes shifts are similar leading to a red-shift of the Tc emission of about 60 nm compared to its rhenium congener. It has also been reported that the luminescence can be electromodulated, which is an important consideration in the designing of a spectroelectrochemical sensor, as electromodulation of luminescence can be effectively used to eliminate interference due to possible contaminant species (Kirchhoff et al. 1986, Chatterjee et al. 2011b, 2012).

Figure 11

Absorption (dashed line) and luminescence (solid line) spectra of [Tc(dmpe)₃]²⁺ (red) in aqueous solution compared with [Re(dmpe)₃]²⁺(blue) in aqueous solution. Adapted with permission from “Del Negro, A.S.; Seliskar, C.J.; Heineman, W.R.; Hightower, S.E.; Bryan, S.A.; Sullivan, B.P. J. Am. Chem. Soc. 2006, 128, 16494.” Copyright © 2006, American Chemical Society.

The similarity in the photophysical properties of the [Tc(dmpe)₃]^2+/+ system with the [Re(dmpe)₃]^2+/+ couple has motivated the use of the latter as a surrogate for the former (Chatterjee et al. 2011b). In aqueous solution, [Re(dmpe)₃]⁺ absorbs at 250 nm (ε=21700 m^-1 cm^-1) but has no absorption in the visible region; [Re(dmpe)₃]²⁺ also absorbs at 250 nm (ε=7740 m^-1 cm^-1) and has visible absorption at 530 nm (ε=2110 m^-1 cm^-1) (Del Negro et al. 2006). In aqueous solution, [Re(dmpe)₃]²⁺ luminesces at 590 nm; [Re(dmpe)₃]⁺ is non-emissive under the same conditions. Additionally, the redox behavior of [Re(dmpe)₃]^2+/+ is very similar to the Tc system, [Re(dmpe)₃]⁺ having a well-defined, electrochemically reversible redox couple for the [Re(dmpe)₃]^+/2+ at E°′=50 mV (vs. Ag/AgCl in water) as shown in Eq. (4) below: (Libson et al. 1988, Roodt et al. 1991, Deng et al. 1997, Kirchhoff et al. 1997)

A model spectroelectrochemical sensor was developed for [Re(dmpe)₃]⁺ (Morris et al. 2009). The sensor consisted of an optically transparent ITO electrode coated with a thin film of various polymers that contacted the sample solution and that were chosen to preconcentrate the analyte. The polymers chosen were: (a) polyacrylic acid (PAA) in a PVA matrix (PAA/PVA), (b) Nafion in a PVA matrix (Nafion/PVA), (c) Nafion coated directly onto the surface of the OTE, and (d) SSEBS, also with no matrix (Gao and Seliskar 1998, Morris et al. 2009). Preconcentration of [Re(dmpe)₃]⁺ was best in the SSEBS film, indicated by a 20-fold increase in peak current compared to a bare OTE after 1 h of continuous cycling in aqueous [Re(dmpe)₃]⁺ solution. The electrochemical and optical absorption properties of [Re(dmpe)₃]^+/2+ contained in SSEBS films were similar to those observed in aqueous solution, indicating that the film did not interfere with these important properties used for detection. Cyclic voltammograms still exhibited the characteristic [Re(dmpe)₃]^+/2+ redox process at 50 mV vs. Ag/AgCl, which is similar to that observed for the respective redox couples in water. The optical absorbances are also similar, with [Re(dmpe)₃]²⁺ absorbing at 530 nm and [Re(dmpe)₃]⁺ being colorless.

A noteworthy observation was the quenching of [Re(dmpe)₃]²⁺ luminescence at 650 nm (when excited at 532 nm) in contact with the SSEBS film. This is presumably a consequence of the highly oxidizing excited state potential of [Re(dmpe)₃]²⁺, which ranks among the highest of the compounds known thus far (Del Negro and Sullivan 2006; Del Negro et al. 2006). Therefore, while [Re(dmpe)₃]⁺ in nonaqueous solution remained stable for a period of several weeks, nonaqueous [Re(dmpe)₃]⁺/SSEBS solutions are immediately oxidized to [Re(dmpe)₃]²⁺ and photoexcitation of these [Re(dmpe)₃]^+/2+/SSEBS solutions results in a loss of any absorbance at 532 nm. While the quenching of luminescence prevented the luminescence electromodulation, the absorption of the compound at SSEBS/ITO electrodes could be electrochemically modulated as shown in Figure 12A. This allowed absorption spectroelectrochemical detection of [Re(dmpe)₃]⁺ at concentrations down to 2×10^-6 m, by reversibly oxidizing the complex from [Re(dmpe)₃]⁺ to [Re(dmpe)₃]²⁺ and monitoring change in absorbance (ΔA) at 530 nm by attenuated total reflectance (ATR) as shown in Figure 12B (Morris et al. 2009).

Figure 12

(A) Optical modulation of absorbance at 530 nm of 0.04 m [Re(dmpe)₃]⁺, 0.1 m KNO₃ in polystyrene-block-poly(ethylene-ran-butylene)-block-polystyrene (SSEBS) for preconcentration times of: (a) 10 min; (b) 1 h; and (c) 2 h; (B) the change in absorbance as a function of [Re(dmpe)₃]⁺ concentration; calibration obtained with 10 min preconcentration time. Reprinted with permission from “Morris, L.K.; Seliskar, C.J.; Heineman, W.R.; Del Negro, A.S.; Bryan, S.A. Electroanal. 2009, 21, 2091.” Copyright © 2009 WILEY-VCH Verlag Gmbh & Co.KGaA, Weinheim.

Spectroelectrochemical sensor for [Tc(dmpe)₃]⁺

While both [Re(dmpe)₃]²⁺ and [Tc(dmpe)₃]²⁺ exhibit emission at longer wavelengths {[Re(dmpe)₃]²⁺, 590 nm; [Tc(dmpe)₃]²⁺, 660 nm}, the emission of [Re(dmpe)₃]²⁺ is rapidly quenched in SSEBS or Nafion films, preventing the emission electromodulation of [Re(dmpe)₃]^+/2+ in these films. However, the emission of [Tc(dmpe)₃]²⁺ in SSEBS is sufficiently long-lived to monitor the [Tc(dmpe)]^2+/+ interconversion by luminescence spectroscopy. Therefore, luminescence detection based on the [Tc(dmpe)₃]^2+/+ electromodulation of the sensor can be efficiently used as a detection mode (Chatterjee et al. 2011b). The sensor consists of an optically transparent ITO electrode coated with a thin film of SSEBS in contact with the sample solution for effective preconcentration.

Spectroelectrochemical modulation of [Tc(dmpe)₃]^2+/+ in the SSEBS film can be accomplished by stepping the potential from -300 mV, at which no luminescence was observed from [Tc(dmpe)₃]⁺, to +700 mV, at which luminescence of [Tc(dmpe)₃]²⁺ was observed. The degree of electromodulation is complete after a short period of time, as indicated by the strong change in luminescence signal on reduction to the Tc(I) complex. Modulation of the luminescence signal is the critical procedure in the spectroelectrochemical sensor concept, which relies on measurement of the change in luminescence as a measure of concentration and to distinguish the target species from other interfering species that might be present in the sample. It is worth noting that there is no significant loss of luminescence intensity of the 660 nm band of [Tc(dmpe)₃]²⁺ during cycling of the electromodulation process for several cycles (shown in Figure 13), which suggests chemical and electrochemical stability of [Tc(dmpe)₃]^2+/+ within the film, and is critical for the durability of the sensor. This sensor allows for the spectroelectrochemical detection over a wide range of solution concentrations, with a detection limit of 24 nm, as shown by the calibration curve shown in Figure 14 (Chatterjee et al. 2011b).

Figure 13

Optical modulation profile (left) of emission intensity (λ_ex=532 nm) upon reversible interconversion of [Tc(dmpe)₃]⁺ (non-emissive) and [Tc(dmpe)₃]²⁺ (emissive) on the application of varying step potential plotted vs. wavelength and time; and varying step potential (right, top) and time of modulation of emission intensity (right, bottom) at 600 nm as a function of applied potential. Reprinted with permission from “Chatterjee, S.; Del Negro, A.S.; Edwards, M.K.; Bryan, S.A.; Kaval, N.; Pantelic, N.; Morris, L.K.; Heineman, W.R.; Seliskar, C.J. Anal. Chem. 2011, 83, 1766.” Copyright © 2011, American Chemical Society.

Figure 14

Spectroelectrochemical sensor calibration curve for the luminescence detection of [Tc(dmpe)₃]²⁺ (in molarity) with preconcentration in a polystyrene-block-poly(ethylene-ran-butylene)-block-polystyrene (SSEBS) film. Reprinted with permission from “Chatterjee, S.; Del Negro, A.S.; Edwards, M.K.; Bryan, S.A.; Kaval, N.; Pantelic, N.; Morris, L.K.; Heineman, W.R.; Seliskar, C.J. Anal. Chem. 2011, 83, 1766.” Copyright © 2011, American Chemical Society.

Effect of concentration of supporting electrolyte on spectroelectrochemical sensor

A practical demonstration of the TcO₄^- sensor on standards and samples from the vadose zone and subsurface water at a radiological site requires: (a) evaluating the effect of the concentration of other soluble components on the sensor response to a target analyte and (b) demonstrating that a spectroelectrochemical sensor can be used directly to measure a target analyte in natural water (e.g., a sample from United States Department of Energy site at Hanford, WA, USA without needing to add supporting electrolyte).

For an ideal sensor performance, it is critical that the presence of soluble components and ions and their concentrations do not interfere with, or adversely affect, the sensor performance. To evaluate this, the effect of varied ionic strength on a spectroelectrochemical sensor has been studied using [Ru(bpy)₃]²⁺ as the model analyte with a sensor consisting of an ITO electrode coated with a thin film of either SSEBS or Nafion. KNO₃, NaNO₃ and Ca(NO₃)₂, in concentrations from 0.1 mm to 1 m were used as representative competing ions in the sample (Morris et al. 2011). The concentration of supporting electrolyte was found to have a significant effect on the performance of the sensor using the model analyte [Ru(bpy)₃]²⁺, reflected by the optical modulations for samples of [Ru(bpy)₃]²⁺ with different concentrations of KNO₃ at a SSEBS/ITO electrode (Figure 15). The ΔA dramatically decreases at higher concentrations of KNO₃, as the K⁺ cation competes with [Ru(bpy)₃]²⁺ for partitioning into the SSEBS film. For both SSEBS and Nafion films, at low supporting electrolyte concentrations, solution resistance is high, resulting in reduced electrochemical and spectroelectrochemical signals. At high supporting electrolyte concentrations, supporting electrolyte cations compete with analytes for ion exchange sites in the film, which reduces the signal. This effect is more pronounced for SSEBS than for Nafion, suggesting the former to be more useful for applications under low ionic strength.

Figure 15

Optical modulation of 0.1 mm [Ru(bpy)₃]²⁺ solutions in the presence of varying supporting electrolyte (KNO₃) concentrations at: (A) polystyrene-block-poly(ethylene-ran-butylene)-block-polystyrene (SSEBS)/ITO electrodes after 30 min of exposure to analyte solution. Comparison of change in absorbance for 0.1 mm [Ru(bpy)₃]²⁺ at Nafion and SSEBS electrodes (B) vs. supporting electrolyte concentration. Reprinted with permission from “Morris, L.K.; Abu, E.A.; Bowman, C.; Estridge, C.E.; Andria, S.E.; Seliskar, C.J.; Heineman, W.R. Electroanal. 2011, 23, 939.” Copyright © 2011 WILEY-VCH Verlag Gmbh & Co.KGaA, Weinheim.

At high concentrations of supporting electrolyte, both films exhibit some loss of sensitivity due to competition of other cations in solution, however, Nafion was far less affected by higher concentrations of competing supporting electrolyte ions than was SSEBS. This is probably true as well for other cations similar to [Ru(bpy)₃]²⁺, a large, relatively hydrophobic organic cation with an extremely large partition coefficient into Nafion relative to H⁺ and other smaller, inorganic cations. A strategy to improve the signal response in solutions with very low ionic strength is to use a slower potential scan rate for modulation. Peak currents increased and peak potential separation decreased by changing the scan rate from 10 mV/s to 5 mV/s when 0.1 mm Ca(NO₃)₂ was used as the supporting electrolyte. The concentration of non-analyte ions in solution and the types of ions present, both non-analyte and analyte ions, and scan rate, can affect both the electrochemical and spectroelectrochemical response of the sensor.

It has been observed that the magnitude of preconcentration is highly dependent on the nature of the polymer polyelectrolye, and also the nature of the supporting electrolyte. Thus, while the preconcentration of [Re(dmpe)₃]⁺ in Nafion and SSEBS films is comparable in the presence of KNO₃ as supporting electrolyte (0.1 m), almost no preconcentration in Nafion is observed in the presence of an equimolar amount of KCl. Moreover, preconcentration in SSEBS in the presence of KBF₄ as supporting electrolyte is almost twice that in the presence of KCl or KNO₃, which are comparable to each other.

Assessing spectroelectrochemical sensor performance in natural and treated water

While the presence of electrolytes in sample solutions is essential for the effective functioning of the electrochemical component of the sensor, for practical applications, it is desirable that the ions already present in the sample soup are sufficient to generate current high enough for detection, without having to add reagents from the outside. This has been evaluated by studying performance of the sensor module in samples of natural (well and river) and treated (tap) water without the addition of supporting electrolyte using [Ru(bpy)₃]²⁺ as a model probe. The studies showed that the absorbance-based spectroelectrochemical sensor can successfully be used to detect [Ru(bpy)₃]²⁺ in these samples, without the need for supporting electrolyte. As a representative example, three water samples were chosen to examine the behavior of the sensor in natural uncontaminated well water (taken from the Hanford nuclear site, in WA, USA) (calibration plot shown in Figure 16), river water (taken from a suburb of Seattle, WA, USA) and treated tap water (from the laboratory in University of Cincinnati, OH, USA). The effect of sample matrix on calibration curves and the ability to quantify the analyte with a standard addition method were examined (Abu et al. 2012). Variations in water hardness and conductivity among these samples had little effect on the calibration curves and detection limits. Thus, the sensor can be used directly on water samples of these types. Detection limits in the low nanomolar range (108 to 264 nm) were achieved. However, these values are specific for [Ru(bpy)₃]²⁺ and are expected to improve for an analyte that undergoes a larger change in molar absorptivity (Δε) with the application of an electrochemical potential. The detection limits are also expected to improve for luminescence detection modalities, which can be much more sensitive than absorption. Sensor performance can also be improved by using ion-exchange films with faster analyte-uptake capacities that would result in shorter analysis times and better regeneration properties. The major outcome of this study was demonstrating the ability to use the spectroelectrochemical sensor on representative water samples, without needing to add supporting electrolyte. This is a significant step forward in the practical application of the sensor to situations where it would be difficult to add reagents to the sample. For example, the sensor could be deployed in a well or a river for continuous monitoring.

Figure 16

Standard addition plot for the detection of [Ru(bpy)₃]²⁺ in Hanford, WA well water; the change in absorbance of [Ru(bpy)₃]²⁺ plotted as a function of [Ru(bpy)₃]²⁺ concentration. [ΔA=(0.203±0.009)×{[Ru(bpy)₃]²⁺}+0.016±0.005], (R²=0.996). Reprinted with permission from “Abu, E.A.; Bryan, S.A.; Seliskar, C.J.; Heineman, W.R. Electroanal. 2012, 24, 1517.” Copyright © 2012 WILEY-VCH Verlag Gmbh & Co.KGaA, Weinheim.

Conclusions

This review demonstrates the ability of the three-component sensor module to effectively and efficiently detect analytes present in natural samples at low limits of detection. It also highlights the ability of the sensor module to detect analytes of interest in the presence of various interfering species in the environmental samples and illustrates the rationale behind the choice of the sensor module for the effective detection of Tc in the environment.