To facilitate the development of molten salt reactor technologies, a fundamental understanding of the physical and chemical properties of molten salts under the combined conditions of high temperature and intense radiation fields is necessary. Optical spectroscopic (UV–Vis–near IR) and electrochemical techniques are powerful analytical tools to probe molecular structure, speciation, thermodynamics, and kinetics of solution dynamics. Here, we report the design and fabrication of three custom-made apparatus: (i) a multi-port spectroelectrochemical furnace equipped with optical spectroscopic and electrochemical instrumentation, (ii) a high-temperature cell holder for time-resolved optical detection of radiolytic transients in molten salts, and (iii) a miniaturized spectroscopy furnace for the investigation of steady-state electron beam effects on molten salt speciation and composition by optical spectroscopy. Initial results obtained with the spectroelectrochemical furnace (i) and high-temperature cell holder (ii) are reported.

There is growing interest in the development of molten salt nuclear reactors (MSRs) in which the fissile material is dissolved in molten salt that also serves as the primary heat transfer fluid. MSRs have a number of safety advantages over contemporary water- and gas-cooled reactor technologies, such as low-pressure operation, a negative temperature coefficient of reactivity, and intrinsically safe shutdown capability.1,2 The soundness of the MSR concept was demonstrated by the Molten Salt Reactor Experiment (MSRE) at Oak Ridge National Laboratory in the 1960s.3,4 However, the MSRE reported several challenges pertaining to the molten salt chemistry that need to be addressed. Additionally, there has not been any practical experience with MSRs since the MSRE, leading to several significant fundamental knowledge gaps that must be bridged for efficient design and innovation of future MSRs, particularly in understanding how their radiation-driven chemistry evolves as fission and corrosion products accumulate in the molten salt (MS). Consequently, understanding how the oxidation and coordination states of solute metal ions (actinides, lanthanides, and transition metals) are influenced by the composition of the base salt matrix (typically mixtures of alkali and alkaline earth halides) is critical. This challenge was made manifest by prior studies showing that radiation generates reactive transients in simple molten salt matrices.5–8 In other systems, such radiolytic transients are capable of driving metal ion oxidation state distributions, degrading more complex molecules, and inducing corrosion processes, thereby affecting the physical and chemical properties of the system.9–12 Consequently, elucidating metal ion redox behavior in high temperature MS radiation environments has real-world consequences in terms of managing metal ion speciation and solubilities to avoid reactor fouling, contamination, and corrosion.

To this end, optical and x-ray spectroscopy tools have been employed to identify metal ion oxidation states and coordination environments and time-resolved electron pulse radiolysis techniques have been used to examine the mechanisms of radiation-induced chemistry in MS media.13–16 However, experiments with MS require specific design considerations due to the high temperatures involved, coupled with the extreme sensitivity of salt chemistry to atmospheric contamination (moisture, oxygen, and other trace impurities), and materials compatibility issues caused by the corrosive nature of MS systems.17 UV–Vis–near-infrared (NIR) spectroscopy has previously been applied to MS systems, and much knowledge about the structure of these high-temperature solutions has been gained through the application of ligand field theory.18,19 Early experimental apparatuses employed for the collection of UV–Vis–NIR spectra were designed to fit within the available space of a spectrophotometer, necessitating complex custom glassware for handling hygroscopic and air-sensitive salts in the open laboratory environment.20–22 The reactivity and corrosivity of MS, particularly in the presence of trace amounts of moisture, limits the useful lifetime of quartz, thus making these experiments time consuming, tedious, and expensive to perform. Advances in high temperature fiber optics have allowed for the separation of the experimental apparatus from the spectrophotometer, reducing the complexity of the furnace required. Recent MS furnace designs for UV–Vis spectroscopy by Pacific Northwest National Laboratory (PNNL)23 and the Korean Atomic Energy Research Institute (KAERI)24 have utilized fiber optic-coupled spectrophotometers with custom furnaces to record spectra of MS solutions relevant to the pyrochemical processing of used nuclear fuel but are limited to a single sample in the furnace at a time. These setups also suffer from wavelength limitations, as a single set of fiber optics is not able to accommodate the entire wavelength range from UV (∼200 nm) through NIR (>1000 nm). Furthermore, additional complexities are introduced when designing MS apparatus for use at beamlines and/or other radiation sources.

In light of these limitations, we report the design and fabrication of three custom-made MS apparatus: (i) a multi-port spectroelectrochemical furnace equipped with optical spectroscopic and electrochemical instrumentation for metal ion spectroscopy and electrochemistry, (ii) a high-temperature cell holder for time-resolved optical detection of radiolytic transients and measurement of their reaction kinetics in MS media, and (iii) a miniaturized spectroscopy furnace for the investigation of steady-state electron beam effects on molten salt speciation and composition using optical spectroscopy. The initial results obtained with the spectroelectrochemical furnace (i) and high-temperature cell holder (ii) are also included in this study. The combination of these custom MS capabilities facilitates the determination of baseline (thermodynamic) metal ion speciation (i.e., oxidation state and coordination environment) and their interaction (kinetics) with key radiolytic transients. The design and fabrication of the third capability are also discussed. Chloride salt matrices are the focus of this study. Although the MSRE operated with mixtures of fluoride salts, chloride salts have attracted interest from several reactor designers due to their performance in a fast neutron spectrum.25 Chlorides have additional advantages compared to fluorides, including higher actinide solubilities and lower melting point solutions formed with actinide trichlorides.25,26 Furthermore, chloride salts are more compatible with quartz-based containment vessels, facilitating ease of interrogation by optical spectroscopies. The equipment described here supports the use of standard or customized commercial quartz spectrophotometer cells.

This MS furnace was designed to allow incorporation of several spectroscopic and electrochemical techniques while accommodating multiple MS samples simultaneously to minimize downtime and allow for direct comparison of sample spectra to a baseline sample.

As shown in Fig. 1, the furnace centers around a multi-cell holder that accommodates five standard 1 × 1 cm2 pathlength quartz cuvettes [Fig. 1(c)]. Conceptually, the furnace can be split into three main sub-assemblies: (i) the furnace core, which is the inner structure of the furnace, (ii) the sample rack that holds up to five MS sample cells and aligns them with optical/beam ports, and (iii) the casement that encloses the furnace and contains the heater and insulation.

FIG. 1.

Spectroelectrochemical molten salt furnace. (a) Cross section view of the Autodesk Inventor model of the furnace (1) handle, (2) electrode assembly, (3) sample rack alignment plate, (4) top insulation, (5) furnace casement, (6) heat shield assembly, (7) furnace core, (8) hole for optic insertion, (9) incident beam optic tube, (10) bottom insulation, (11) graphite sample holder, (12) shutter tube, (13) hole for heater lead, (14) shutter handle, (15) thermocouple insertion hole, (16) electrode alignment hole, (17) sample rack alignment pin, (18) threaded rod for sample rack alignment, (19) quartz cuvette, and (20) receiving optic tube (aluminosilicate fiber insulation and heater coil not shown for clarity). (b) Side view of the assembled furnace inside the glovebox. (c) Sample rack inside the glovebox with molten salt samples and electrode assembly in place. (d) Shutter assembly. (e) Three electrode assembly.

FIG. 1.

Spectroelectrochemical molten salt furnace. (a) Cross section view of the Autodesk Inventor model of the furnace (1) handle, (2) electrode assembly, (3) sample rack alignment plate, (4) top insulation, (5) furnace casement, (6) heat shield assembly, (7) furnace core, (8) hole for optic insertion, (9) incident beam optic tube, (10) bottom insulation, (11) graphite sample holder, (12) shutter tube, (13) hole for heater lead, (14) shutter handle, (15) thermocouple insertion hole, (16) electrode alignment hole, (17) sample rack alignment pin, (18) threaded rod for sample rack alignment, (19) quartz cuvette, and (20) receiving optic tube (aluminosilicate fiber insulation and heater coil not shown for clarity). (b) Side view of the assembled furnace inside the glovebox. (c) Sample rack inside the glovebox with molten salt samples and electrode assembly in place. (d) Shutter assembly. (e) Three electrode assembly.

Close modal

The furnace core [Fig. 1(a)] is a 3.75 in diameter × 0.25 in wall × 7 in tall stainless-steel 316 (SS316) tube (McMaster Carr, 89495K5), machined to precisely locate optic/beam ports in all angular and linear dimensions. Optic/beam ports were machined from 1 in diameter. SS316 rod (McMaster Carr, 89325K24). To maintain precise alignment of the optics over a broad temperature range (room temperature to 800 °C), the core and optic/beam ports were first rough machined to within 0.020 in of final dimensions and annealed according to ASTM A269 Section 1.2 at 845 °C for 1 h, removing any residual stresses from machining or forming.27 The surfaces were then machined to final tolerances and the threaded holes are tapped. The outer end of the beam ports were also tapped to make a female 0.25 in national pipe tapered (NPT) thread on the inner bore to accommodate the fiber optic assemblies. Due to the radial symmetry of the core, the majority of the movement from thermal expansion upon heating is radially outward, minimizing the loss in angular alignment of the collimating optics for incident and receiving optical fibers. The large thermal mass of the core, combined with the insulation and heat shielding, gives excellent thermal stability despite the heater being placed below the furnace. The bolt pattern on the faces of the mounting flanges allows for the installation of custom optic/beam ports or blank plates, as required, at each location. For example, spectrophotometric experiments employ bespoke UV–Vis–NIR optical fibers (Fiberguide Industries, CB24133-X), Fig. 1(b), that are connected to the furnace by optic tubes designed to allow complementary 0.25-in male national pipe tapered (NPT) fittings to be attached and removed without disassembly of the entire furnace. The assembly of all threaded components was performed with high-temperature MoS2-based anti-seize to prevent galling (Bostik Never-Seez® Blue Moly). Finally, to provide temperature feedback to the furnace controller, a hole was drilled and tapped for a female 0.125-in NPT fitting at a height 1 cm below the final level of the sample locating block, allowing insertion of a dual-zone K-type thermocouple (Omega, SCASS-062G-6-DUAL, not shown). A 0.125-in diameter, 12-in long K-type thermocouple (McMaster Carr, 39095K64) is inserted through the top of the sample holder to measure the salt temperature, as shown in Fig. 1(b). This thermocouple is located ∼0.3 cm from the cuvette wall and can be inserted directly next to the sample of interest for temperature monitoring and calibration.

The sample rack [Fig. 1(c)] is comprised of a graphite sample locating block machined from stock on hand, threaded rods (McMaster Carr, 90575A149) to adjust the relative height of the samples to the optic/beam-ports, a welded series of heat shields and spacers, a locating plate, and a handle assembly. The locating plate has five small locating holes to allow positive and repeatable alignment of the samples with the furnace core when the sample rack is rotated. This gives the ability to quickly switch between samples and/or analytical techniques by a simple rotation of the rack to the correct orientation, without removing the samples from the furnace or re-verification of the sample alignment relative to the optics. Due to the need to collect the transmitted beam for spectrophotometric measurements, the optical path opposite each MS sample cuvette was kept clear, necessitating an odd number of positions in the sample rack in pentagonal geometry, allowing for up to five samples to be placed in the furnace simultaneously. A cylindrical tube with five pairs of opposing apertures [Fig. 1(d)] rotates around its long axis to serve as a shutter that allows the optic/beam ports to be protected from direct exposure to salt vapors, while signal is not being recorded.

The casement [Figs. 1(a) and 1(b)] is a SS316 enclosure repurposed from a glove-box circulation pump (MBRAUN), with machined holes to allow for passage of the fiber optics. It houses a 3 in diameter spiral FeCrAl heating element (MHI, Inc., MC-GAXP-375) for resistive heating of the furnace from below. The heater is protected from potential salt spills by a 3.5 in diameter × 0.0625 in thick quartz disc (McMaster Carr, 1357T49). The core, quartz disc, and heater coil rest in a piece of calcium silicate insulation (McMaster Carr, 9353K33) machined to prevent direct contact between the heater and quartz plate and to maintain relative placement of the furnace core within the furnace casement. The diameter of the quartz plate is such that the inner diameter of the furnace core is covered, preventing spills from reaching the heating element as well as shorting between the core and the heating element. An electrical junction box below the casement contains the electrical connections for the heater coil and provides strain relief for these connections.

Initial testing of the furnace resulted in multiple failures of the heater coil after relatively short periods of use, resulting in development efforts to extend coil lifetime. The root cause of these failures was traced to binding of the heater coil caused by thermal expansion, resulting in short-circuiting across the coil and localized melting. Consequently, it was found that the connection between the coil leads and the 12-gauge copper wire used for power transmission must be made flexible to allow for coil movement without binding. Silver solder was used to attach a 12-gauge flexible braided cable to the coil directly, which was then crimped to the copper conductors. Flexible fiberglass sheathing was used for high temperature electrical insulation for these connections. Additionally, custom spacers to prevent direct contact between adjacent coils were made by spreading three 0.5 in wide lines of Aremco Ceramabond 569 paste radially from the center of the coil to the edge, spaced approximately every 120° around the coil after the rest of the bottom parts were assembled. The thickness of the Ceramabond layer was kept as thin as possible by spreading with a gloved finger pressed firmly on the coil. After curing in air at 200 °C for 2 h using the coil as a heat source, the result was a series of ceramic spacers that prevent contact between adjacent coils. As the Ceramabond did not adhere firmly to the coil nor to the ceramic baseplate, movement of the coil is not encumbered by the presence of the spacers. No subsequent failures of the heater coil have occurred following the implementation of these modifications; however, due to the challenges associated with the use of the flat spiral heating coil, future revisions may benefit from redesign of the heating system.

Temperature control and power for the furnace are provided by a custom-made furnace controller located external to the glovebox. A Watlow proportional–integral-derivative (PID) controller (part number PM61RFJ-AAAAAAA) regulates the pulse width of a silicon-controlled rectifier (SCR) (Watlow, part number DA1024FO00) via a 4 mA–20 mA signal, which then modulates the output side of a high current transformer. The transformer converts 120 V power to the 63 V power required by the heater coil. Maximum current at 100% duty cycle is 27 A. An overtemperature limit controller (Watlow part number PM6L1AJ-AAAAAAA) monitors the second zone of the aforementioned K-type thermocouple to ensure fail-safe operation should any part of the furnace control circuit fail in a manner that results in an extended temperature excursion. The limit controller operates a relay that controls the power inputs for both the transformer and PID controller. Power is transmitted to the heater coil via a custom high-power feedthrough using 12-gauge cabling and Amphenol three prong 16A connectors (part numbers 97-3101A-20-19-S and 97-3101A-20-19-P) with the cables potted into a KF40 flange (Douglas Electrical Components, custom).

Currently installed analytical capabilities include transmission UV–Vis and Vis–NIR spectrophotometry measured by using a fiber-coupled Agilent Cary 5000 UV–Vis–NIR spectrometer along with electrochemical measurements performed via electrode assemblies inserted through the heat shield from the top of the sample rack. An array of KF40 flanges were added to the side of the inert glovebox housing to allow for feedthrough of all fiber-optics, electrode leads, thermocouple wires, and furnace power connections. There are total ten beam ports (five pairs), out of which four (two pairs) are currently used for the UV–Vis and Vis–near IR fibers. Having three additional pairs of optical ports will enable the development of fiber-coupled Raman spectroscopy and other capabilities in the future.

For spectrophotometry, fiber-optics were made to specification by Fiberguide Industries. Two sets of fibers are used, identical in construction other than their wavelength range. One set is used for UV–Vis spectra between 200 nm and 1200 nm and the other covers the Vis–NIR range between 350 nm and 2300 nm. To maximize signal integrity, the 0.22 numerical aperture 400 μm core multi-mode fibers connect between the spectrophotometer (equipped with a Harrick FiberMate 1 fiber optic coupler) and the optical port of the furnace through a KF40 flange permanently attached and hermetically sealed to the fiber jacketing and stainless-steel armor. Changing the set of fibers in use (UV–Vis or Vis–NIR) is accomplished by manually changing the set of fibers connected to the spectrophotometer. Approximately 18 in of the fiber ends are gold clad to allow operation at temperatures up to 700 °C. Optic/beam-port alignment is through a 0.25 in Swagelok to male 0.25 in NPT fitting that attaches to the fiber end and threads into the optic port mounted to the furnace core. This allows the fibers to be removed from the furnace for maintenance or for higher-temperature operation if necessary. A small quartz plano–convex lens is mounted on a custom stainless-steel collimator assembly to collimate the signal. The collimator lens assembly protects the fiber end from being etched by the salt vapors and is easily replaced if degradation of performance due to salt vapor exposure is observed.

For electrochemical measurements, electrode assemblies are created by inserting 1 mm diameter rods of the material of choice into one of the 1.5 mm bores in a four-bore alumina tube with an outer diameter of 0.2 in (CoorsTek) Fig. 1(e). This allows for a typical three-electrode arrangement consisting of a quasi-reference electrode (QRE), a working electrode, and a counter electrode, as well as other electrode arrays if desired.

Pulse radiolysis transient absorption spectroscopy is a technique that uses short pulses of radiation to initiate chemical reactions that are observed by time-resolved measurement of optical absorbance in the UV, visible, NIR, and/or mid-infrared ranges. It permits the detection of short-lived, radiation-induced transient species and measurement of their reaction rates. The Laser Electron Accelerator Facility (LEAF)28 at Brookhaven National Laboratory (BNL) is based on a photocathode electron gun accelerator that delivers 7 ps–60 ps electron pulses to three experimental beamlines (designated A, B, and C) having (A) ultrafast laser-time-of-flight stroboscopic detection,29 (B) UV–Vis–NIR digitizer-based detection,28 and (C) mid-IR digitizer-based detection.30 To support our pulse radiolysis investigations of MS solutions, the sample holder for the LEAF B-line digitizer-based target station was redesigned as a set of modular parts that can be configured for experiments at high temperatures, cryogenic temperatures, or with secondary containment for radioactive samples, as shown in Fig. 2.28 

FIG. 2.

High-temperature molten salt cell holder for picosecond pulse radiolysis (insulation removed for clarity). (a) Partial section view (cut-out) of the modular sample block (1) rotatable flange, (2) forward and rear water-cooled mirror sections, (3) mirror cube with front silvered mirror, (4) sample compartment, insulation not shown, (5) quartz cuvette, (6) water cooled aluminum heat sink, (7) rear mounting plate, and (8) Faraday cup assembly. (b) Photograph of the high-temperature setup mounted on the LEAF B-line: (1) beamline flange, (2) rotatable flange, (3) forward water-cooled mirror section, (4) water-cooled aluminum heat sink mounted to the high-temperature sample compartment, (5) cuvette, (6) Faraday cup, (7) Faraday cup signal cable, (8) thermocouple connection, (9) cartridge heater connection, and (10) cooling lines. The rear water-cooled mirror section between parts 4 and 6 is obscured by the cooling lines. (c) Detailed view the of high temperature sample compartment (near side insulation not shown for clarity): (1) quartz cuvette, (2) sample compartment, (3) water cooled aluminum heat sink, (4) ¼ in-20 tapped hole, (5) alignment hole, (6) 304 stainless steel cuvette holder sections (right side cut away for clarity), (7) alumina bushing, (8) alumina rod, (9) thermocouple, (10) spring, (11) cartridge heater, and (12) insulation.

FIG. 2.

High-temperature molten salt cell holder for picosecond pulse radiolysis (insulation removed for clarity). (a) Partial section view (cut-out) of the modular sample block (1) rotatable flange, (2) forward and rear water-cooled mirror sections, (3) mirror cube with front silvered mirror, (4) sample compartment, insulation not shown, (5) quartz cuvette, (6) water cooled aluminum heat sink, (7) rear mounting plate, and (8) Faraday cup assembly. (b) Photograph of the high-temperature setup mounted on the LEAF B-line: (1) beamline flange, (2) rotatable flange, (3) forward water-cooled mirror section, (4) water-cooled aluminum heat sink mounted to the high-temperature sample compartment, (5) cuvette, (6) Faraday cup, (7) Faraday cup signal cable, (8) thermocouple connection, (9) cartridge heater connection, and (10) cooling lines. The rear water-cooled mirror section between parts 4 and 6 is obscured by the cooling lines. (c) Detailed view the of high temperature sample compartment (near side insulation not shown for clarity): (1) quartz cuvette, (2) sample compartment, (3) water cooled aluminum heat sink, (4) ¼ in-20 tapped hole, (5) alignment hole, (6) 304 stainless steel cuvette holder sections (right side cut away for clarity), (7) alumina bushing, (8) alumina rod, (9) thermocouple, (10) spring, (11) cartridge heater, and (12) insulation.

Close modal

The modular sample block [Fig. 2(a)] consists of an interchangeable sample compartment (1.6 in wide) in the middle and two sections on the upstream and downstream ends (with respect to the electron beam) with mirrors that transmit the analyzing light anti-colinearly to the electron beam. The analyzing light enters the cell holder via the mirror next to the Faraday cup (part 8) and exits by the mirror next to the rotating flange (part 1). The anti-colinear alignment minimizes the collection of the Cerenkov light produced by the electron beam traversing the sample because the Cerenkov light is radiated in a cone projected in the forward direction of the electron beam. The holder is designed in such a way that sample compartments can be changed in minutes while preserving electron beam, probe light, and detector alignment. Apertures in the sample block align the electron beam and probe light. The sample block is mounted directly to the end of the LEAF B beamline [Fig. 2(b)] using a two-part rotatable flange that permits adjustment to ensure that the path of the probe light is horizontal through the sample block and forward to the detector. The flange measures 3.25 in square and 0.375 in thick and it has through-holes for bolts and holes for alignment pins. Once the flange is aligned, the block can be removed in seconds and remounted without losing light beam alignment on the detector. The two mirror sections on either side of the sample compartment hold cubic mirror mounts and double as water-cooled heat sinks. First-surface silvered mirrors (0.5 × 0.75 × 0.020 in3) are inserted into the two mirror cubes and held with setscrews [Fig. 2(a), inset showing part 3]. One mirror cube is then pressed into each mirror section and held by friction. The mirrors dissipate heat by conduction but are slowly damaged by the intense radiant energy. Mirrors are easily replaced when needed. Alignment pins are pressed through both mirror sections and the sample compartment to make the assembly kinematic. Water channels are used to keep the sample block assembly at ambient temperature and protect the beamline’s vacuum window from excessive heat.

A Faraday cup assembly is attached to the rear mounting plate on the downstream end of the sample block, as defined by the path of the electron beam. The purpose of the Faraday cup is to collect the portion of the high-energy electron beam that completely traverses the sample so that a shot-by-shot quantification of the radiolytic dose can be recorded. The 3 × 3 in2 assembly consists of an aluminum standoff block 1.5 in thick mounted to the mirror section and a 0.25 in thick aluminum backplate. The interior of the standoff block has a circular hole with a diameter of 1.5 in Within that cavity, a copper cylinder of 1.25 in diameter and 1.125 in long is mounted to the backplate using nylon screws for electrical isolation and a 0.25 in thick, 1.25 in diameter Teflon disc for support. The copper cylinder is used to collect the remaining incident electron beam. It is wired to the center pin of a female BNC bulkhead connector whose shell is grounded to the backplate. A 330 pF capacitor inside the block connects the BNC center pin to ground on the backplate. The standoff block is chamfered on one side to allow space for the capacitor. In operation, the Faraday cup is connected to a low-noise current preamplifier with adjustable gain (Stanford Research Systems model SR570) to produce a signal proportional to the charge of the electron pulse that traversed the sample, which can be integrated and recorded as a relative measure of the radiolytic dose. Chemical dosimetry is used during experimental runs to convert the Faraday cup measurements to absolute radiation doses.

The high-temperature sample compartment is designed to hold 5 × 5 mm2 Spectrosil® quartz cuvettes that have been customized with Teflon valves (Starna Cells, Inc.), as shown in Fig. 2(b). The cuvette holder is made in two parts of 304 stainless steel [Fig. 2(c)]. The dimensions of the sample area are 1.2 in wide × 0.9 in deep × 1.7 in high. Each side has a 100 W, 120 VAC cartridge heater (McMaster-Carr 4877K111) with a maximum rated temperature of 760 °C. Temperature control is achieved using an Omega Engineering CN7800 series temperature controller. A K-type thermocouple is located between one of the cartridge heaters and the cuvette. Alumina rods suspend the two sides of the cuvette holder within the sample compartment assembly. One side of the cuvette holder is spring loaded to improve thermal contact, while the other is fixed to maintain optical alignment. Alumina silica ceramic fiber (McMaster-Carr 93285K18-Lydall Performance Materials, Inc.) insulation is cut and packed around the cuvette holder. Two water-cooled aluminum heat sinks (custom) are bolted to the sample area section and are supplied with 20 °C water from a chiller during high temperature experiments. The mirror sections are also supplied in parallel by the same chiller. With this configuration, we have successfully melted MgCl2 (m.p. 715 °C) for pulse radiolysis experiments, but we cannot melt KCl (m.p. 770 °C). Future, higher-temperature work will require custom-made cartridge heaters made from Kanthal® A-1 wire that can operate up to 1400 °C, although the practical operating limit for the cell holder may depend on other factors. For the studies that we anticipate doing, 900 °C should be sufficient. The modified cartridge heaters will be controlled by an Omega CND3 temperature controller with linear voltage output and connected to a BK 9103 360-W power supply.

The modular nature of the cell holder assembly permits rapid changeover for different kinds of pulse radiolysis experiments in addition to the high-temperature setup, including more conventional experiments between 0 °C and 90 °C, working with radioactive samples and cryogenic experiments. The sample compartment for standard pulse radiolysis experiments is U-shaped with a 1.1 in wide × 2.1 in deep rectangular cavity [Fig. 3(a)]. It accommodates a variety of sample configurations by simply changing the cuvette holder insert. Each brass holder insert has a front and rear aperture just smaller than the internal width of the cuvette to keep the probe light from bypassing the sample and to ensure overlap of the electron and probe light beams. One such insert consists of a water/propylene glycol-jacketed brass block for variable-temperature measurements between 0 °C and 90 °C.

FIG. 3.

Modular sample compartments for various experimental conditions. (a) Standard configuration (partial section view): (1) cuvette, (2) brass cuvette holder, (3) sample compartment, (4) alignment hole, and (5) ¼ in-20 taped hole. (b) Partial section view of the assembled two-piece cell holder insert for secondary containment of radioactive samples: (1) thumb screw, (2) upper cuvette container, (3) lower cuvette container, (4) quartz window, (5) 1 ml catch area, (6) actinide sample cuvette, (7) sample compartment, (8) alignment hole, and (9) ¼ in-20 tapped hole. (c) Cross section of the cryogenic sample compartment assembly (partial section view): (1) cuvette, (2) cooling gas, (3) cartridge heater, (4) thermocouple, (5) brass cuvette holder, (6) dry nitrogen purge channel, (7) 3D-printed PLA plastic outer shell, (8) alignment hole, (9) ¼ in-20 tapped hole, and (10) sample compartment.

FIG. 3.

Modular sample compartments for various experimental conditions. (a) Standard configuration (partial section view): (1) cuvette, (2) brass cuvette holder, (3) sample compartment, (4) alignment hole, and (5) ¼ in-20 taped hole. (b) Partial section view of the assembled two-piece cell holder insert for secondary containment of radioactive samples: (1) thumb screw, (2) upper cuvette container, (3) lower cuvette container, (4) quartz window, (5) 1 ml catch area, (6) actinide sample cuvette, (7) sample compartment, (8) alignment hole, and (9) ¼ in-20 tapped hole. (c) Cross section of the cryogenic sample compartment assembly (partial section view): (1) cuvette, (2) cooling gas, (3) cartridge heater, (4) thermocouple, (5) brass cuvette holder, (6) dry nitrogen purge channel, (7) 3D-printed PLA plastic outer shell, (8) alignment hole, (9) ¼ in-20 tapped hole, and (10) sample compartment.

Close modal

Another insert is designed to provide secondary containment for radioactive samples sealed in commercial thread-capped quartz cuvettes Fig. 3(b). It consists of two sections (upper and lower) of 3D-printed PLA plastic designed to protect the cuvettes from damage during handling and to contain any leakage that may occur to avoid contamination of the experimental systems. Two circular S1-UV grade fused silica windows (12.7 mm diameter, 1 mm thick, Esco Optics P105040) are used to seal the apertures for the electron and optical beams.

A low temperature sample compartment that can cool samples down to −196 °C was also developed Fig. 3(c). The temperature is controlled by an Omega CN7800 temperature controller and a K-type thermocouple. Cooling is achieved with nitrogen blow-off from a liquid nitrogen Dewar. The temperature controller activates a solenoid valve and allows cold gas to flow through a brass cuvette holder. Generally, the heater is not used for low temperature work but aids in warming the cell and going to elevated temperatures. The brass cuvette holder sits in a 3D-printed PLA plastic holder that has channels for buffer gas. Dry nitrogen is fed up through the plastic holder at 5 standard cubic feet per minute, which flows past the mirrors. This prevents cuvette and mirror fogging at temperatures below −80 °C. Water at 20 °C is also run through the mirror sections to prevent condensation on the beamline window and the Faraday cup, which could induce false dosimetry readings.

To investigate the steady-state effects of low linear energy transfer (LET) ionizing radiation on MS matrices, a miniaturized version of the spectroelectrochemical furnace described above in Sec. II A was designed by INL and installed on the end of the Notre Dame Radiation Laboratory (NDRL) 2.8 MeV Van de Graaff accelerator beamline, as shown in Fig. 4. Electron radiolysis results in the formation of transient species due to ionization of the MS, as detailed in the pulsed radiolysis experiment section. Electrons so produced can reduce any available metal ions leading to different coordination environments from the parent medium, which can then be detected by UV–Vis spectroscopy. The formation of an observable amount of reduced species will require high radiation doses that can be acquired through long term steady-state radiolysis with an intense electron beam.

FIG. 4.

Electron beam irradiation molten salt furnace: (a) cross section view of the Autodesk Inventor model of the furnace showing (1) the top alignment plate, (2) locating pins, (3) heat shields, (4) threaded rods for sample suspension, (5) furnace core, (6) machined firebrick insulation, (7) standard quartz cuvette, (8) optic tube, (9) cuvette holder, (10) heater base, and (11) hole for electrical lead to heater. (b) Complete assembly with furnace core, insulation, heater coil, and cuvette is shown. (c) Sample rack removed from furnace, showing the insertion of the flame sealed cuvette.

FIG. 4.

Electron beam irradiation molten salt furnace: (a) cross section view of the Autodesk Inventor model of the furnace showing (1) the top alignment plate, (2) locating pins, (3) heat shields, (4) threaded rods for sample suspension, (5) furnace core, (6) machined firebrick insulation, (7) standard quartz cuvette, (8) optic tube, (9) cuvette holder, (10) heater base, and (11) hole for electrical lead to heater. (b) Complete assembly with furnace core, insulation, heater coil, and cuvette is shown. (c) Sample rack removed from furnace, showing the insertion of the flame sealed cuvette.

Close modal

The electron beam irradiation furnace is designed to house and align a single quartz cuvette with perpendicular optic/beam-ports: one pathway allows for the passage of the incident 2.8 MeV electron beam through the sample, while the second pathway facilitates the measurement of spectroscopic data collected in transmission mode through a fiber-optic design identical to that detailed in Sec. II A. Here, an Ocean Optics Flame spectrometer equipped with an inline light attenuator is used to collect spectra of the salt samples. Thermal insulation for the electron beam irradiation furnace consists of two calcium silicate firebricks machined to be a negative impression of the furnace core and optics, as well as external ceramic fiber insulation (not shown). Due to space constraints, this MS furnace capability operates outside of an inert glovebox. Consequently, all MS samples were prepared and characterized at INL and flame sealed under vacuum in bespoke cuvettes to maintain inert conditions.

This section details the sample preparation and initial results for MS capabilities (A) and (B). The MS capability (C) is a miniature version (four ports instead of ten) of the capability (A) and will be operated in similar manner for electron beam irradiation experiments at NDRL—with one set of ports used for irradiation and the other for the optical spectroscopy measurements. The furnace (C) has been installed and tested at NDRL; however, the initial results using this capability are not available to be included in this study.

Chromium trichloride (CrCl3) was chosen as an analyte for testing the performance of these apparatus since it is a prevalent corrosion product in MS systems and has two soluble oxidation states (II and III) accessible electrochemically with UV–Vis absorption spectra in MS media.31,32 Lithium chloride–potassium chloride (LiCl–KCl) eutectic salt was used as the solvent salt matrix as it has a relatively low melting point of 352 °C while being stable with minimal vaporization at much higher temperatures, it is transparent in the spectral range available for study and has been well investigated in the open literature, allowing for direct comparison to the current work.31,33–35 Anhydrous salts, CrCl3 and LiCl–KCl eutectic, were obtained from Sigma Aldrich at 99.99% purity and sealed under argon. No further purification procedures for the salts were performed.

To ensure a thorough understanding of the Cr system prior to irradiation studies, baseline measurements of the relevant properties needed to be established. To this end, the spectroelectrochemical MS furnace was used to record optical and electrochemical measurements for trivalent chromium (Cr3+) dissolved in LiCl–KCl eutectic. Samples were prepared by weighing the required masses of CrCl3 and LiCl–KCl eutectic directly into UV-grade quartz cuvettes (FireflySci). The samples were then loaded into the furnace and allowed to equilibrate at 500 °C for 1 h prior to immersion of the electrodes and collection of spectra.

For electrochemical measurements, electrode assemblies were created by inserting 1 mm diameter wires into 1.5 mm bores in a four-bore alumina tube with an outer diameter of 0.2 in (CoorsTek). The working and counter electrodes were tungsten (Alfa Aesar) and the quasi-reference electrode (QRE) was nickel (Alfa Aesar). The scale of the system demanded use of a minimally invasive reference electrode; therefore, a QRE (Ni wire) was utilized. QREs are commonly employed for high-temperature molten salt systems and do not drift significantly in short-duration use.36,37 Due to the reported use and adequate stability of Ni QRE a detailed assessment of RE stability was not performed during the test. The same electrode assembly was used for all electrochemical measurements over the course of the experiment. The four-bore ceramic tube allows for two-, three-, and four-electrode configurations as desired. A Bio-Logic VSP potentiostat was interfaced with a laptop computer running EC-Lab software. To establish an electrochemical potential window, initial cyclic voltammograms were recorded in blank LiCl–KCl eutectic at 500 °C (not shown). The same batch of blank LiCl–KCl eutectic also served as the baseline sample for spectroscopic measurements. Spectra were recorded using an Agilent Cary 5000 UV–Vis–NIR spectrophotometer. The wavelength scale of the spectrophotometer over the range of 880 nm to 280 nm was verified and calibrated using a NIST-certified reference material comprising an aqueous solution of didymium perchlorate at room temperature permanently sealed by heat fusion in a high-quality far-UV quartz cell (Starna Scientific, certificate number 72641). The LiCl–KCl eutectic solution was then doped with 0.1 wt. % CrCl3 via direct addition of a weighed amount of CrCl3 to solid LiCl–KCl eutectic. No mixing was performed, and the solution was allowed to equilibrate for 2 h at 500 °C before the absorption spectrum of Cr3+ was recorded [curve 1, Fig. 5(a)]. Electrodes were introduced into the solution for electrochemical measurements and the corresponding cyclic voltammogram was recorded, as shown in Fig. 5(b), demonstrating a two-step reduction of Cr3+, first to Cr2+ and then to Cr0. The relationship and position of these redox peaks are reasonable when compared to results reported by Cotarta et al.38 From this cyclic voltammogram, the potential for potentiostatic polarization for bulk electrolysis was selected in order to partially reduce Cr3+ to Cr2+ without undergoing complete reduction to Cr0. A constant potential of −0.5 V vs Ni QRE was then applied for 1800 s and the current monitored, as shown in the chronoamperometric curve in Fig. 5(c).

FIG. 5.

Spectroelectrochemical measurements for Cr3+ in a LiCl–KCl eutectic matrix at 500 °C: (a) UV–Vis spectroscopic monitoring of the bulk electrolytic reduction of Cr3+ to Cr2+ induced by potentiostatic polarization at −0.5 V vs Ni QRE (spectrum 1 is before polarization, and spectra 2–5 indicate increased reduction time), (b) cyclic voltammogram for Cr3+ ↔ Cr2+ and Cr2+ ↔ Cr0 reactions, and (c) chronoamperometric curve for the Cr3+ → Cr2+ reduction, performed concurrently with the spectra recorded in (a).

FIG. 5.

Spectroelectrochemical measurements for Cr3+ in a LiCl–KCl eutectic matrix at 500 °C: (a) UV–Vis spectroscopic monitoring of the bulk electrolytic reduction of Cr3+ to Cr2+ induced by potentiostatic polarization at −0.5 V vs Ni QRE (spectrum 1 is before polarization, and spectra 2–5 indicate increased reduction time), (b) cyclic voltammogram for Cr3+ ↔ Cr2+ and Cr2+ ↔ Cr0 reactions, and (c) chronoamperometric curve for the Cr3+ → Cr2+ reduction, performed concurrently with the spectra recorded in (a).

Close modal

While potentiostatic polarization was being performed, UV–Vis absorption spectra were simultaneously recorded every minute over the course of one hour to monitor the extent of reduction. A selection of these spectra, displayed in Fig. 5(a), show a clear decrease in the peak intensity of the 4F(4T1) and 4F(4T2) modes of Cr3+ (peaks at 526 nm and 833 nm, respectively) and the emergence of the 5E2 mode of Cr2+ around 1000 nm.31,39 Furthermore, there is an apparent shift of the charge transfer band to higher energies as the reduction proceeds, which is consistent with the previously reported spectra of the Cr3+ and Cr2+ species.32 These initial results successfully demonstrate the capability of the spectroelectrochemical MS furnace and the techniques used to drive and probe different oxidation states of species in MS.

The solvated electron (es) is a powerful reducing agent and one of the key transient species arising from irradiation of MS.5–8,11 In aqueous solutions, es is efficiently scavenged by Cr3+,

es+Cr3+Cr2+kwater=4.0×1010M1s1(24).
(1)

Furthermore, the electrochemical results above show that Cr3+ is easily reduced to the stable divalent state, Cr2+, in molten LiCl–KCl eutectic. Consequently, the reactivity profile of es will strongly influence the chemical evolution of MSR fuels.

To demonstrate the efficacy of the high-temperature pulse radiolysis cell holder, es reaction kinetics with Cr3+ were measured in a 5 mM CrCl3 solution in LiCl–KCl eutectic (42 mol. % KCl) at 50 °C intervals between 400 °C and 650 °C. As above, LiCl–KCl eutectic and CrCl3 were used without further purification and subsequently manipulated in an Ar-atmosphere glove box (MBRAUN UNIlab). Sample mixtures were prepared by weighing out the constituents in custom 5 × 5 mm2 Spectrosil Quartz Starna cuvettes fitted with in-line valves (Fig. 2). Prior to experimentation, the cuvettes were evacuated by a dry scroll vacuum pump (XDS5C, BOC Edwards) to avoid cell rupture from overpressure upon heating. The sample cuvette was placed in the cell holder and allowed to thermally equilibrate for at least 10 min at the desired temperature before measurement. To verify that thermal equilibration had been attained, kinetics measurements were performed over several minutes during which the observed es decay rate coefficients remained consistent. Reaction kinetics, induced by the ∼50 ps-long, 8.7 MeV electron pulses from the LEAF accelerator, were measured by time-resolved absorption spectroscopy using a pulsed 75 W Xe lamp, a FND-100 silicon photodiode detector, and interference filters (∼10 nm bandpass) for wavelength selection of the analyzing light. The photodiode signals were digitized with a 600 MHz, 12-bit oscilloscope (LeCroy WaveRunner HRO 66Zi). Dosimetry for this experiment was based on the yield of the (SCN)2–• radical dimer anion in N2O-saturated, 10 mM KSCN aqueous solution (measured at 470 nm, Gε = 2.6 × 10–4 m2 J−1),40 corrected for the electron density ratio between water (3.34 × 1023e/cm3) and LiCl–KCl eutectic (4.82 × 1023e/cm3). The dose per pulse was 20 Gy–27 Gy.

The reaction of es with Cr3+ was observed as a function of temperature by following the decay of the es absorbance at its band maximum of 671 nm, as shown in Fig. 6. The es is formed promptly within the detection system response time of ∼2 ns, and it decays exponentially through reaction with Cr3+. (The consequent formation of the Cr2+ product cannot be observed at 671 nm under these conditions. We observe in Fig. 5 and Ref. 32 that the reduction of Cr3+ to Cr2+ would result in a negative transient absorbance signal (bleach) at 671 nm, but due to the small extinction coefficient differences between the two Cr valence states (<100 M−1 cm−1), this signal would be very small compared to the massive and broad transient absorbance signal of the solvated electron, ε ≈ 15 000 M−1 cm−1 to 20 000 M−1 cm−1.) The observed lifetime of the es under these conditions was ∼20 ns at 400 °C. For comparison, the es lifetime at 400 °C in the same batch of LiCl–KCl eutectic without added CrCl3 is ∼135 ns (data not shown), indicating that background es decay contributed ∼15% to the observed es decay rate in the presence of 5 mM Cr3+. The observed rate of es decay increased with increasing temperature. We also observe that the peak absorbance at the end of the signal rise time decreased with increasing temperature. We attribute this to increasing rates of very fast processes that scavenge electrons within the signal rise time, including electron–hole recombination and pre-solvated electron scavenging.41 

FIG. 6.

Decay kinetics at 671 nm for the es from pulsed electron irradiation of 5 mM CrCl3 in LiCl–KCl eutectic at 400 °C (black filled square), 450 °C (red filled circle), 500 °C (blue filled triangle), 550 °C (green filled inverted triangle), 600 °C (light blue filled diamond), and 650 °C (purple filled hexagon). Inset: Arrhenius plot for the reaction of the es with Cr3+ calculated from first-order fits to the decay kinetics.

FIG. 6.

Decay kinetics at 671 nm for the es from pulsed electron irradiation of 5 mM CrCl3 in LiCl–KCl eutectic at 400 °C (black filled square), 450 °C (red filled circle), 500 °C (blue filled triangle), 550 °C (green filled inverted triangle), 600 °C (light blue filled diamond), and 650 °C (purple filled hexagon). Inset: Arrhenius plot for the reaction of the es with Cr3+ calculated from first-order fits to the decay kinetics.

Close modal

We estimate that the concentration of the es produced in Fig. 6 is of the order of 6 μM–15 μM, assuming that the extinction coefficient at 671 nm is in the range of 15 000 M−1 cm−1 to 20 000 M−1 cm−1. This assumption is based on the measured es absorbance produced by electrolysis in KCl at 800 °C42 and is consistent with es extinction coefficients in ionic liquids43,44 and molecular solvents.45 In this case, the kinetics for reaction (1) are assumed to be pseudo-first-order as the concentration of Cr3+ is much higher than that of the es. First-order fits to the es decay curves as a function of temperature afforded observed rate coefficients of (0.6–2.2) × 108 s−1. Using these values, the corresponding Arrhenius plot (Fig. 5, inset) can be constructed to calculate an activation energy (Ea) of 31 ± 2 kJ mol−1. The estimated second-order rate coefficient for reaction (1) at 400 °C is 1.2 × 1010 M−1 s−1. A more rigorous, Cr3+ concentration-dependent redetermination of the rate constant and activation parameters of this reaction, and a detailed spectroscopic study using higher doses and spanning from the UV to the NIR to detect the Cr2+ product, will be part of a larger subsequent study. These initial results demonstrate the capability of the high temperature MS cell-holder capability to conveniently measure reaction kinetics for radiation-induced transients in MS at high temperatures.

Versatile high-temperature spectroscopic, electrochemical, and radiolysis capabilities have been designed and fabricated for the study of fundamental MS properties. The described spectroelectrochemical furnace and high-temperature pulse radiolysis cell holder are designed to allow flexibility with regard to sample types and temperatures studied while maintaining inert atmosphere conditions. Initial results demonstrate their ability to collect high-quality data on MS. Specifically, the electrolytic reduction of Cr3+ to Cr2+ in LiCl–KCl was monitored over time by UV–Vis–NIR spectrophotometry. The 4F(4T1) and 4F(4T2) modes of Cr3+ were observed prior to reduction, and the 5E2 mode of Cr2+ was observed following partial reduction. The kinetics of the reaction of es with Cr3+ in LiCl–KCl was easily measured at temperatures up to 650 °C. While only the absorption of the es was detected in this example, transient metal ion species with extinction coefficients of ≥1000 M−1 cm−1 are also amenable to detection. A broad range of studies are planned to combine these experimental capabilities with steady-state irradiations, molecular dynamics simulations, synchrotron x-ray absorption and scattering techniques (XAFS, XANES, and differential PDF), and corrosion experiments in order to construct a thorough and predictive understanding of metal ion speciation and reactivity in MS at high temperature and under irradiation in MSR environments.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

This work was supported as part of the Molten Salts in Extreme Environments Energy Frontier Research Center, funded by the U.S. Department of Energy (US-DOE), Office of Science, Basic Energy Sciences, at BNL under Contract No. DE-SC0012704, at INL under Contract No. DE-AC07-05ID14517, and at Notre Dame under subcontract to BNL. For the support of furnace design and fabrication work at INL, R.G. acknowledges the Laboratory Directed Research and Development (LDRD) Program under DOE Idaho Operations Office Contract No. DE-AC07-05ID14517 and a subcontract under LDRD (INL) to University of Nevada, Reno via Award No. 204471 (UNR Award No. AWD-01-00001719). The Laser Electron Accelerator Facility of the BNL Accelerator Center for Energy Research is supported by the U.S. DOE Office of Basic Energy Sciences, Division of Chemical Sciences, Geosciences, and Biosciences under Contract No. DE-SC0012704. The radiolysis facilities of the Notre Dame Radiation Laboratory (NDRL) are supported by the U.S. DOE, Office of Basic Energy Sciences, Division of Chemical Sciences, Geosciences, and Biosciences, through Grant No. DE-FC02-04ER15533.

The authors gratefully acknowledge Thomas Morgan, Ronald Wallace, and Dean Burt at INL and Joe Admave, Kiva Ford, and Phil George at NDRL for their expertise in machining, fabrication, and electrical control systems, without which the furnaces could not have been constructed. We also thank Steven M. Frank at INL for his technical guidance.

1.
D. T.
Ingersoll
,
Small Modular Reactors: Nuclear Power Fad or Future?
(
Woodhead Publishing
,
2015
).
2.
M. D.
Carelli
and
D. T.
Ingersoll
,
Handbook of Small Modular Nuclear Reactors
(
Elsevier
,
2014
).
3.
P. N.
Haubenreich
and
J. R.
Engel
, “
Experience with the molten-salt reactor experiment
,”
Nucl. Appl. Technol.
8
(
2
),
118
136
(
1970
).
4.
W. R.
Grimes
, “
Chemical research and development for molten-salt breeder reactors
,” ORNL-TM-1853,
1967
.
5.
I. E.
Makarov
,
T. N.
Zhukova
,
A. K.
Pikaev
, and
V. I.
Spitsyn
, “
Oxidizing agents produced by radiolysis of alkali-metal halide melts
,”
Bull. Acad. Sci. USSR, Div. Chem. Sci.
31
(
4
),
662
666
(
1982
).
6.
A. K.
Pikaev
,
I. E.
Makarov
, and
T. N.
Zhukova
, “
Solvated electron in irradiated melts of alkaline halides
,”
Radiat. Phys. Chem.
19
(
5
),
377
387
(
1982
).
7.
I. E.
Makarov
,
T. N.
Zhukova
, and
A. K.
Pikaev
, “
Transient species in radiolysis of melted alkali metal halides
,”
Radiat. Eff.
22
(
1
),
71
72
(
1974
).
8.
H.
Hagiwara
,
S.
Sawamura
,
T.
Sumiyoshi
, and
M.
Katayama
, “
Pulse radiolysis study of transient species in LiCl-KCl melt
,”
Int. J. Radiat. Appl. Instrum., Part C
30
(
2
),
141
144
(
1987
).
9.
R. A.
Peterson
 et al, “
Review of the scientific understanding of radioactive waste at the U.S. DOE Hanford site
,”
Environ. Sci. Technol.
52
,
381
396
(
2018
).
10.
B. B. J.
Mincher
and
S. P.
Mezyk
, “
Radiation chemical effects on radiochemistry: A review of examples important to nuclear power
,”
Radiochim. Acta
97
,
519
534
(
2009
).
11.
G. V.
Buxton
,
C. L.
Greenstock
,
W. P.
Helman
, and
A. B.
Ross
, “
Critical review of rate constants for reactions of hydrated electrons, hydrogen atoms and hydroxyl radicals (OH/O) in aqueous solution
,”
J. Phys. Chem. Ref. Data
17
(
2
),
513
886
(
1988
).
12.
B. J.
Mincher
and
J. F.
Wishart
, “
The radiation chemistry of ionic liquids: A review
,”
Solvent Extr. Ion Exch.
32
(
6
),
563
583
(
2014
).
13.
J.
Brynestad
,
C. R.
Boston
, and
G. P.
Smith
, “
Electronic spectra and coordination of nickel centers in liquid lithium chloride–potassium chloride mixtures
,”
J. Chem. Phys.
47
(
9
),
3179
3189
(
1967
).
14.
A. K.
Plkaev
,
B. G.
Ershov
, and
I. E.
Makarov
, “
Influence of the nature of a matrix on the reactivity of electrons in irradiated systems
,”
J. Phys. Chem.
79
(
26
),
3025
3034
(
1975
).
15.
G. P.
Smith
,
D. W.
James
, and
C. R.
Boston
, “
Optical spectra of TI+, Pb2+, and Bi3+ in the molten lithium chloride–potassium chloride eutectic
,”
J. Chem. Phys.
42
(
6
),
2249
2250
(
1965
).
16.
A.
Merwin
,
W. C.
Phillips
,
M. A.
Williamson
,
J. L.
Willit
,
P. N.
Motsegood
, and
D.
Chidambaram
, “
Presence of Li clusters in molten LiCl-Li
,”
Sci. Rep.
6
,
25435
(
2016
).
17.
F.
Lantelme
and
H.
Groult
,
Molten Salts Chemistry
(
Elsevier, Inc.
,
2013
).
18.
K. E.
Johnson
, “
Ligand field stabilization in molten salts: Spectrochemical and nephelauxetic series
,”
Electrochim. Acta
11
(
2
),
129
(
1966
).
19.
M.
Hatanaka
and
S.
Yabushita
, “
Mechanisms of f–f hypersensitive transition intensities of lanthanide trihalide molecules: A spin–orbit configuration interaction study
,”
Theor. Chem. Acc.
133
(
8
),
1517
(
2014
).
20.
D. M.
Gruen
and
R. L.
Mcbeth
, “
Oxidation states and complex ions of uranium in fused chlorides and nitrates
,”
J. Inorg. Nucl. Chem.
9
,
290
301
(
1959
).
21.
B. R.
Sundheim
and
J.
Greenberg
, “
High temperature modification of the Beckman DU spectrophotometer
,”
Rev. Sci. Instrum.
27
(
9
),
703
704
(
1956
).
22.
J.
Greenberg
and
L. J.
Hallgren
, “
Techniques for measuring the infrared absorption spectra of fused salts
,”
Rev. Sci. Instrum.
31
(
4
),
444
445
(
1960
).
23.
C. A.
Schroll
,
S.
Chatterjee
,
T. G.
Levitskaia
,
W. R.
Heineman
, and
S. A.
Bryan
, “
Electrochemistry and spectroelectrochemistry of europium(III) chloride in 3LiCl–2KCl from 643 to 1123 K
,”
Anal. Chem.
85
,
9924
9931
(
2013
).
24.
B. Y.
Kim
,
D. H.
Lee
,
J.-Y.
Lee
, and
J.-I.
Yun
, “
Electrochemical and spectroscopic investigations of Tb(III) in molten LiCl–KCl eutectic at high temperature
,”
Electrochem. Commun.
12
(
8
),
1005
1008
(
2010
).
25.
M.
Taube
and
J.
Ligou
, “
Molten chlorides for fast breeder reactor problems and possibilities
,” Eidg. Institut fur Reaktorforschung Wureniingen Schweiz Report No. EIR-Bericht Nr.215,
1972
.
26.
O.
Beneš
and
R. J. M.
Konings
,
Molten Salt Reactor Fuel and Coolant
(
Elsevier
,
2012
), Vol. 3.
27.
Standard Specification for Seamless and Welded Austenitic Stainless Steel Tubing for General Service, ASTM A269.
28.
J. F.
Wishart
,
A. R.
Cook
, and
J. R.
Miller
, “
The LEAF picosecond pulse radiolysis facility at Brookhaven National Laboratory
,”
Rev. Sci. Instrum.
75
(
11
),
4359
4366
(
2004
).
29.
A. R.
Cook
and
Y.
Shen
, “
Optical fiber-based single-shot picosecond transient absorption spectroscopy
,”
Rev. Sci. Instrum.
80
(
7
),
073106
(
2009
).
30.
D. C.
Grills
,
J. A.
Farrington
,
B. H.
Layne
,
J. M.
Preses
,
H. J.
Bernstein
, and
J. F.
Wishart
, “
Development of nanosecond time-resolved infrared detection at the LEAF pulse radiolysis facility
,”
Rev. Sci. Instrum.
86
(
4
),
044102
(
2015
).
31.
D. M.
Gruen
and
R. L.
McBeth
, “
The coordination chemistry of 3D transition metal ions in fused salt solutions
,”
Pure Appl. Chem.
6
(
1
),
23
48
(
1963
).
32.
M. V.
Smirnov
and
A. M.
Potapov
, “
Redox potentials and electronic absorption spectra of dilute solutions of nickel and chromium chlorides in molten alkali chlorides
,”
Electrochim. Acta
39
(
1
),
143
149
(
1994
).
33.
A. S.
Basin
,
A. B.
Kaplun
,
A. B.
Meshalkin
, and
N. F.
Uvarov
, “
The LiCl–KCl binary system
,”
Russ. J. Inorg. Chem.
53
(
9
),
1509
1511
(
2008
).
34.
H.-C.
Eun
,
H.-C.
Yang
,
Y.-J.
Cho
,
H.-S.
Park
,
E.-H.
Kim
, and
I.-T.
Kim
, “
Separation of pure LiCl-KCl eutectic salt from a mixture of LiCl-KCl eutectic salt and rare-earth precipitates by vacuum distillation
,”
J. Nucl. Sci. Technol.
44
(
10
),
1295
1300
(
2007
).
35.
B. Y.
Kim
and
J.-I.
Yun
, “
Temperature effect on fluorescence and UV-vis absorption spectroscopic properties of Dy(III) in molten LiCl–KCl eutectic salt
,”
J. Lumin.
132
(
11
),
3066
3071
(
2012
).
36.
A. J.
Bard
and
L. R.
Faulkner
,
Electrochemical Methods: Fundamentals and Applications
(
John Wiley & Sons, Inc.
,
2001
).
37.
M. R.
Shaltry
,
R. O.
Hoover
, and
G. L.
Fredrickson
,
J. Electrochem. Soc.
167
(
11
),
116502
(
2020
).
38.
A.
Cotarta
,
J.
Bouteillon
, and
J. C.
Poignet
, “
Electrochemistry of molten LiCl–KCl–CrCl3 and LiCl–KCl–CrCl2 mixtures
,”
J. Appl. Electrochem.
27
(
6
),
651
658
(
1997
).
39.
G. P.
Smith
, “
Review of electronic absorption spectra of molten salts
,” ORNL-3411,
1963
.
40.
G. V.
Buxton
and
C. R.
Stuart
, “
Re-evaluation of the thiocyanate dosimeter for pulse radiolysis
,”
J. Chem. Soc. Faraday Trans.
91
(
2
),
279
281
(
1995
).
41.
J. F.
Wishart
,
A. M.
Funston
,
T.
Szreder
,
A. R.
Cook
, and
M.
Gohdo
, “
Electron solvation dynamics and reactivity in ionic liquids observed by picosecond radiolysis techniques
,”
Faraday Discuss.
154
,
353
363
(
2012
).
42.
D.
Nattland
,
T.
Rauch
, and
W.
Freyland
, “
Electron localization in metal–molten salt solutions: An optical study with in situ variation of composition
,”
J. Chem. Phys.
98
(
6
),
4429
4436
(
1993
).
43.
J. F.
Wishart
and
P.
Neta
, “
Spectrum and reactivity of the solvated electron in the ionic liquid methyltributylammonium bis(trifluoromethylsulfonyl)imide
,”
J. Phys. Chem. B
107
(
30
),
7261
7267
(
2003
).
44.
T.
Kondoh
 et al, “
Pulse radiolysis study of ion-species effects on the solvated electron in alkylammonium ionic liquids
,”
Radiat. Phys. Chem.
78
(
12
),
1157
1160
(
2009
).
45.
A. M.
Funston
and
J. F.
Wishart
, “
Dynamics of fast reactions in ionic liquids
,” in
Ionic Liquids III A: Fundamentals, Progress, Challenges, and Opportunities
, edited by
R. D.
Rogers
and
K. R.
Seddon
(
American Chemical Society
,
Washington
,
2005
), Vol. 901, pp.
102
116
.