1. Background

Access to accurate thermophysical property data is vital the in nuclear safety analysis, reactor design, licensing, and operation of advanced molten salt reactors (MSRs). MSRs are considered a potential game-changing technology for next-generation nuclear reactors that could be cost effective, safe, and more sustainable than traditional commercial nuclear power plants. A significant knowledge gap exists in the data for the fundamental properties relevant to fuels and coolants for MSRs which needs to be addressed in order to expedite the technical readiness level of different MSR design concepts. To this end, a proposed fuel salt, NaCl-PuCl₃, and the binary system was investigated in this work.

The authors are aware of only one other experimental study on the NaCl-PuCl₃ system reported by Bjorklund et al. in 1959 in which the melting temperature of PuCl₃ was determined to be 767°C, and the NaCl-PuCl₃ eutectic composition contained 36mol% PuCl₃ with a melting temperature of 453°C [1]. PuCl₃ was reported to be blue green in color with a density of 5.708 g∙cm^‒3 [2].

There are several documented synthesis paths for producing PuCl₃, with differing precursors. PuCl₃ can be synthesized from the reaction of PuO₂ or plutonium (III) oxalate with PCl₅, SCl₂, HCl-H₂, COCl₂, and CCl₄. However, these are either not scalable reactions or do not yield a pure PuCl₃ product. An aqueous synthesis route using plutonium (III) oxalate as the starting material is shown in Equations 1–2 [3].

The difficulties associated with obtaining a pure solution of Pu(III) and removing all water of crystallization complicates an aqueous preparation of PuCl₃; however, despite these difficulties a 98% pure, PuCl₃ has been obtained [3]. The synthesis and purification of PuCl₃ is typically achieved through high temperature pyrochemical processing where the PuO₂, feedstock is converted to Pu metal by direct reduction in CaCl₂ through the multicycle direct oxide reduction (MDCOR) process [4]. Note the Pu metal produced by the MDCOR process is not always of high purity due to inherent impurities in the CaCl₂ and the buildup of ²⁴¹Am by the radioactive decay of ²⁴¹Pu (t_1/2 = 14.29 years). To remove ²⁴¹Am, chlorine gas is sparged through the molten plutonium converting the Pu to PuCl₃ which then reacts with ²⁴¹Am to produce AmCl₃ [5]. However, if no Am is present, this route can be used to produce PuCl₃. Care must be taken during this synthesis path because incomplete chlorination will result in a PuCl₃ product containing small quantities of Pu metal.

Johnson et al. [6] studied the presence of Pu metal in PuCl₃ and its effects on the melting temperature where it was determined that a eutectic transition exists at a composition of 7mol% Pu in PuCl₃ and melts at 740°C. The effect on the NaCl-PuCl₃ binary system when using PuCl₃ containing Pu metal remain undetermined.

2. Experimental

For a detailed analysis of the binary NaCl-PuCl₃ phase diagram, four samples were prepared using high-purity NaCl (99.99, anhydrous), and PuCl₃. The PuCl₃ had been in sealed storage at the at the Idaho National Laboratory as received from Los Alamos National Laboratory in 2010. The NaCl was conditioned by heating to 400°C under vacuum at 3.6 Pa for 4 hours in a distillation furnace to remove possible trace moisture impurities. The NaCl did not experience mass loss after the conditioning in the distillation furnace, indicating the material was of high purity. The NaCl and PuCl₃ salts were handled inside argon-atmosphere gloveboxes where the moisture and oxygen levels were kept below 0.1 ppm and 5 ppm, respectively.

No conditioning or purification was performed on the PuCl₃ prior to use; the material was assumed to be highpurity PuCl₃ but was not verified upon receipt in 2010. The nuclear material transaction report (NMTR) provided the plutonium isotopic breakdown of the PuCl₃ sample as shown in Table 1, and visually, the PuCl₃ appeared to be dark green-blue to grey in color. Using information from the NMTR, the PuCl₃ was calculated to be 94.05wt% PuCl₃ with 5.95wt% Pu-metal present. Since there were no means of PuCl₃ purification at the time, the NaCl-PuCl₃ phase diagram was investigated using the as received PuCl₃ salt.

Table 1

Isotopic composition, in weight percent, of plutonium from the PuCl₃ salt in 2010

Sample preparation consisted of weighing the NaCl and PuCl₃ in required proportions, then mixing the two components in a tapered, glassy carbon crucible, see Figs. 1(a) and Fig. 1(b). In total, six samples were used in this study. The compositions are provided in Table 2; two of the samples were the unary NaCl and PuCl₃ salts. Four samples (S2– S5) were heated in a furnace to 850°C and cooled to an ambient temperature before being removed. All four salt mixtures were easily removed from the glassy carbon crucibles, and crucible-salt interaction was not observed. The salt buttons, see Fig. 1(c), were size reduced Fig. 1(d) using an agate mortar and pestle which was cleaned between samples. Each salt was placed in a corresponding airtight, leak-checked metal container before being transferred for analysis.

Fig. 1

Pictures of the salt blending process. (a) glassy carbon crucibles containing PuCl₃ and NaCl, (b) PuCl₃ and NaCl mixture prior to heating, (c) “button” of NaCl/PuCl₃ mixture after heating, (d) pulverized eutectic button, used for thermal analysis.

Table 2

Compositions of the samples used to study the NaCl-PuCl₃ phase diagram

The blended samples were transferred to the Fresh Fuels Glovebox at MFC for analysis using the simultaneous thermal analyzer (STA) model STA449F3A-0171-M containing a type-S, differential scanning calorimeter-thermogravimetric (DSC/TG) sample carrier. Thermal analysis using the STA is capable of collecting DSC and TG data simultaneously for each sample, when properly calibrated. The STA was operated inside an inert argon glove box using ultra-high-purity argon protective gas and purge gas with a flow rate of 50 and 20 ml∙min^–1, respectively. Non-reactive, unsealed glassy carbon crucibles with glassy carbon lids containing a small hole were used for all standards and samples. Samples were run using a temperature and sensitivity calibration, accurate in the range of 20 to 950°C, generated using five high purity calibration standards with heating rates of 10 and 2°C∙min⁻¹. The melting or polymorphic transition temperature and peak area of each standard were calculated by averaging the onset and peak area for three heating cycles. The accuracy of the temperature calibration curve was verified using a silver standard where the experimental melting temperature was 962.0°C, within ± 1°C of the reported silver melting temperature of 961.78°C [7].

During data collection and sample preparation, the glovebox atmosphere was maintained at less than 5 ppm O₂ and 0.1 ppm moisture. The reported polymorphic, eutectic, and melting-point temperatures were determined by averaging the onset temperature (first peak) and peak temperature (subsequent peaks) from three heating cycles. The averaged transition temperature for each sample was then plotted against the heat rate, and the linear regression was used to remove the effects of thermal lag on the reported temperature value [8].

Initially, a melting-point and stability analysis was performed on the PuCl₃ to gain insight on purity and investigate the effects of Pu metal or possibly radiolysis after being in storage for several decades. Subsequent thermal analysis on the five remaining samples was performed to determine the effects of PuCl₃ concentration on the melt/ transition temperature and provide insights into the stable temperature range for measuring future thermophysical properties.

3. Results & Discussion

3.1 Evaluation of PuCl₃

Fig. 2(a) shows the programmed temperature profile (black, dashed) for the STA along with the TG (blue) curve for the PuCl₃ sample with a heating rate of 10°C∙min⁻¹. In total, the sample experienced four heating and cooling cycles ranging from 650 to 820°C for the first heating cycle then 600 to 820°C for the remaining three cycles. After each cooling segment, the sample was held isothermally for 20 minutes before proceeding with the subsequent heating segment. At the start of the experiment, the relative sample mass was 100%, corresponding to a sample mass of 35.0 mg. During the four heating and cooling cycles, oscillations in the mass curve can be seen due to buoyancy of the sample carrier since the argon gas expands in the furnace causing this type of sample carrier to “float”. However, upon completing the experiment and returning to ambient temperature, the resulting mass gain was 0.27%, or 0.0945 mg, for the PuCl₃ sample. No difference in the mass of the sample and crucible pre- and post- experiment, on a four-place analytical balance, was observed.

Fig. 2

STA results for PuCl₃ with a heating rate of 10°C∙min⁻¹. (a) temperature profile and TG curve, (b) first heating DSC curves: segment 1 (grey) and segment 2 (black), (c) DSC curves for heating cycles 2–4 with the first derivative of heating cycle 2 (grey dashed line), (d) cooling curves, cycles 2–4.

When evaluating the temperature dependent changes in samples, typically the first heating curves are discarded. However, in the case of PuCl₃, inspection of the first heating curve is warranted in order to observe inflections at lower temperatures and gain insight on the purity of the sample. The first heating segment occurring from 20 to 650°C in Fig. 2(b) shows no inflections or energy changes specifically below 200°C, indicating the sample had not been contaminated with moisture.

DSC heating segments used to calculate the melting temperature of the PuCl₃ are shown in Fig. 2(c). From these segments, it can be concluded that the sample did not interact with the glassy carbon crucible; otherwise, a noticeable variation (i.e., shift) in energy as a function of temperature would be observed for each heating cycle. For the PuCl₃ used in this work, the singular broad peak occurring from 715 to 760°C is deemed the melting peak. To determine the onset temperature, the first derivative of the DSC curve is taken and shown for heating in Fig. 2(c). The average onset of the three heating cycles is 714°C, while the average peak-melting temperature was determined to be 749°C. Analysis of the 2°C∙min⁻¹ heating curves yields an average melting onset temperature of 734°C and a peak-melting temperature of approximately 747°C.

The narrow crystallization peak for each cooling cycle is shown in Fig. 2(d). Inspection of the cooling curves indicates that crystallization or solidification occurs over an approximate 12°C peak width, but temperature specific information on PuCl₃ solidification cannot be extracted from the peaks due to supercooling of the salt. The shape and width of the peak on cooling indicate a pure sample with a liquid-to-solid transition and no evidence of a solid solution or solid transition. Cooling curves suggest a pure sample while the broad peak and lower experimental melting temperature extracted from the heating curve raise concerns over the purity of the PuCl₃ and presence of Pu metal.

Integrating peak areas for the heating (Fig. 2(c)) and cooling (Fig. 2(d)) curves of PuCl₃ yield the enthalpy of fusion (Δ H f o) and crystallization (Δ H c o), respectively. The average experimentally determined values are Δ H f o = 102.2 ± 10 J∙g^‒1 (35.8 kJ∙mol^‒1) and Δ H c o = ‒101.4 ± 10 J∙g^‒1 (35.5 kJ∙mol^‒1). Calculated literature reported enthalpy of fusion values includes 55.0 ± 5.0 kJ∙mol^‒1 [9] and 63.6 kJ∙mol^‒1 [10]. The authors are unaware of any experimentally determined values for the enthalpy of fusion or crystallization for PuCl₃, in correlation with the observation of others [11].

There are several reported values for the melting temperature of PuCl₃; the most widely used melting temperature value is 767 ± 2°C [1]. However, others have reported a PuCl₃ melting temperature of approximately 760 ± 5°C [3] and 733 ± 10°C [12]. Another study using purified PuCl₃ determined the material melted completely in the range of 760 to 765°C, and when cooled, crystallization began at 760°C [13]. The PuCl₃ salt used in this study has a melting temperature of 746.5°C approximately 20°C lower than reported by Bjorklund et al. [1]. There are several possibilities, each discussed in detail, for the lower melting temperature observed in this work including: (1) moisture/oxygen contamination, (2) self-radiolysis, and (3) presence of Pu metal.

The simplified, Pu-PuCl₃ phase diagram from Johnson et al. [6], reprinted with permission from Elsevier in Fig. 3, aids in determining the concentration of Pu present in the PuCl₃. The mol% of Pu and PuCl₃ for both possible compositions is shown in Table 3 along with the melting temperature at the composition determined from the Pu- PuCl₃ phase diagram. The melting temperature of PuCl₃ used in this study suggests there is 6.2mol% of Pu metal in the PuCl₃ sample, which very closely corresponds to 5.95wt% Pu provided in the NMTR shipping information. The 6.2mol% Pu in PuCl₃ has a melting temperature of 745.6°C while the 5.95wt% Pu in PuCl₃ has a melting temperature of 746.5°C, matching the melting temperature of the material investigated here. The 7.87mol% Pu composition is deemed unfeasible due to the higher melting point and lack of any high-temperature peaks in the DSC curves shown in Fig. 2(c).

Fig. 3

Composition of the PuCl₃ used in this work () plotted over the focused Pu-PuCl₃ system. Reprinted from [6] with permission from Elsevier.

Table 3

Calculated values for the two possible compositions of the PuCl₃ salt used in this study determined using the Pu-Cl phase diagram and the corresponding melting temperature determined in this work

3.2 Binary Phase Diagram Study

To further investigate the effects of PuCl₃ containing Pu metal on the NaCl-PuCl₃ phase diagram, six compositions on the NaCl-PuCl₃ phase diagram were analyzed on the STA, including samples of PuCl₃ and NaCl. A summary of the DSC curves generated for each sample with a heating rate of 10 and 2°C∙min⁻¹ is shown in Fig. 4. For both the NaCl (Fig. 4(a)) and PuCl₃ (Fig. 4(f)), salt samples with a single melting peak at both heating rates were observed, which is typical for a unary salt. The sample with 83.1mol% NaCl showed only two peaks; the first at 447°C is indicative of the eutectic transition while the second peak at 741°C correlates to the melting/liquidus temperature for that sample.

Fig. 4

DSC heating curves of the six samples used in this NaCl-PuCl₃ phase diagram study in which Pre-Eut. = pre-eutectic peak, Eut. = eutectic peak, and MP = melting peak.

The samples in Figs. 4(c)–4(e) show three peaks, with the final peak being relatively more subtle. The first peak .abeled as the pre-eutectic peak is most likely a solid-to-solid phase transition due to the many crystal structures of Pu metal as well as the lack of reported liquid phase formation below 400°C. Analysis of the data generated from the STA yields the averaged values in Table 4 for the pre-eutectic, eutectic, and liquidus temperature, obtained by taking the average transition temperature for three heating cycles at each heating rate used in this study.

Table 4

Results from DSC curves generated in the STA

The data from Table 4 were plotted against the experimental data generated by Bjorklund et al. [1] and are shown in Fig. 5. Bjorklund et al. purified the samples by bubbling HCl(g), ensuring full chlorination of the PuCl₃. It has been reported that the incomplete conversion of Pu metal, hydride, or oxide to PuCl₃ occurs when using a variety of synthesis paths [3] as seen in this work’s PuCl₃. The resulting NaCl-PuCl₃ phase diagram is of high importance for MSRs, pyroprocessing, fuel storage, and safeguarding material because the presence of unreacted metallic Pu can drastically affect the melting temperature and change the eutectic composition of the NaCl-PuCl₃ binary which can in turn alter operating parameters such as temperature and Pu loading in the salt.

Fig. 5

NaCl-PuCl₃ system developed in this work (black) with data and curve fit lines overlayed with the NaCl-PuCl₃ binary data and trendlines (red) from Bjorklund et al.

4. Conclusion

The melting temperature of the PuCl₃ used in this work, Pu-Cl phase diagram, and shipping information all aided in the conclusion that the PuCl₃ in this study has approximatly 5.9–6.2mol% Pu present. This work highlights the importance of an in-depth understanding and characterization of the starting material, especially when used in nuclear applications utilizing the NaCl-PuCl₃ system. The effects of Pu-metal due to insufficient chlorination or solubility of Pu metal in PuCl₃, must be well understood, as even small concentrations will affect the melting temperature and eutectic composition. There are several important points to extract from the NaCl-PuCl₃ phase diagram study using the PuCl₃ containing Pu metal.