- Email: [email protected]
Melting behavior of mixed U–Pu oxides under oxidizing conditions
Nuclear Instruments and Methods in Physics Research B 374 (2016) 125–128
Contents lists available at ScienceDirect
Nuclear Instruments and Methods in Physics Research B journal homepage: www.elsevier.com/locate/nimb
Melting behavior of mixed U–Pu oxides under oxidizing conditions Michal Strach a,c, Dario Manara b,⇑, Renaud C. Belin a, Jacques Rogez c a
CEA, DEN, DTEC, SECA, LCC, Cadarache F-13108, Saint-Paul-Lez-Durance, France European Commission, Joint Research Centre, Institute for Transuranium Elements, P.O. Box 2340, 76125 Karlsruhe, Germany c IM2NP, UMR CNRS 7334 – Aix Marseille University, Case 251, Avenue Escadrille Normandie Niemen, 13397 Marseille Cedex 20, France b
a r t i c l e
i n f o
Article history: Received 29 May 2015 Received in revised form 16 December 2015 Accepted 4 January 2016 Available online 10 February 2016 Keywords: MOX U–Pu–O Laser melting Oxidation
a b s t r a c t In order to use mixed U–Pu oxide ceramics in present and future nuclear reactors, their physical and chemical properties need to be well determined. The behavior of stoichiometric (U,Pu)O2 compounds is relatively well understood, but the effects of oxygen stoichiometry on the fuel performance and stability are often still obscure. In the present work, a series of laser melting experiments were carried out to determine the impact of an oxidizing atmosphere, and in consequence the departure from a stoichiometric composition on the melting behavior of six mixed uranium plutonium oxides with Pu content ranging from 14 to 62 wt%. The starting materials were disks cut from sintered stoichiometric pellets. For each composition we have performed two laser melting experiments in pressurized air, each consisting of four shots of different duration and intensity. During the experiments we recorded the temperature at the surface of the sample with a pyrometer. Phase transitions were qualitatively identified with the help of a reflected blue laser. The observed phase transitions occur at a systematically lower temperature, the lower the Pu content of the studied sample. It is consistent with the fact that uranium dioxide is easily oxidized at elevated temperatures, forming chemical species rich in oxygen, which melt at a lower temperature and are more volatile. To our knowledge this campaign is a first attempt to quantitatively determine the effect of O/M on the melting temperature of MOX. Ó 2016 Elsevier B.V. All rights reserved.
1. Introduction Recently, numerous studies have been devoted to the reassessment of the melting temperatures of PuO2, UO2 and mixtures of the two compounds (MOX). Main focus has been put on the minimization of the sample-crucible interactions effects [1], which apparently had affected many previous results [2]. It has been concluded, that in the case of actinides (An), the higher the oxygen potential of a given An–O compound, the higher the discrepancies one might expect between results from experiments using some kind of container and container-less ones [3]. Thus, for UO2 the melting temperatures found in older studies and the new data determined using container-less techniques differ only slightly, while for PuO2 (which has a much higher oxygen potential than the aforementioned compound) the difference was found to be up to 300 K. The melting behavior of MOX with different U/Pu ratios has been thoroughly analyzed by Böhler et al. [3], and it was found that a minimum of the melting temperature curve exists between x(PuO2) = 0.4 and x(PuO2) = 0.7. What was also concluded in this study was the apparent effect of the oxygen stoichiometry ⇑ Corresponding author. http://dx.doi.org/10.1016/j.nimb.2016.01.032 0168-583X/Ó 2016 Elsevier B.V. All rights reserved.
(oxygen to metal ratio O/M) on the results. The described laser heating experiments were carried out either under pressurized argon or air, depending on the predominant element, U and Pu respectively. Although the findings indicated that the impact of O/M variations during thermal treatment were slight and did not have a strong impact on the melting behavior in the presented cases, the authors concluded that further investigations are needed to determine the uncertainties concerning the O/M determination. The cited reference provides compelling argumentation in favor of the accuracy of the presented results concerning stoichiometric compounds (O/M = 2.00 at the beginning of the thermal treatment and at the melting point). But one can imagine a scenario, where the fuel experiences strongly oxidizing or reducing atmospheres or environments before or during melting. In this instance, the melting behavior would surely be affected. In another study [4], we found that oxidation of MOX with Pu contents even as high as 62 mol% results in the formation of significant amounts of oxygen-rich phases, such as MO2+x and M3O8, and possible U/Pu differences in the two resulting phases. This is due to the fact that uranium can increase its valence up to +6 and Pu can be incorporated in the resulting structures.
126
M. Strach et al. / Nuclear Instruments and Methods in Physics Research B 374 (2016) 125–128
In the present work we describe a series of experiments aimed at studying the effect of a strong departure from and O/M = 2.00 insitu on the melting behavior of the studied compounds, and on the applied measurement method itself.
2. Materials and methods Six MOX samples were prepared by co-milling, pressing and sintering stoichiometric UO2 and PuO2 powders in appropriate proportions. The fabrication process was detailed in ref. [5]. Disks were cut form the resulting dense pellets, X mm in diameter and 1–2 mm in height. Stoichiometry was achieved by atmosphere control during sintering (Ar + 5%H2 + H2O) and verified by powder XRD measurements. The Pu content was verified by titration and estimated at 14, 24, 35, 46, 54 and 62 mol%. Details concerning the laser heating setup have been listed in ref. [1,6–8]. Surface temperatures of the studied samples were measured by sub-millisecond resolution pyrometry. The uncertainty was estimated at ±50 K. The equipment used for heating the samples beyond the melting point was a TRUMPFÒ Nd:YAG cw laser radiating at 1064.5 nm, which power–time profiles can be programmed with 1 ms resolution. Pulses ranged in duration from 100 to 1000 ms and power from 180 to 675 W. During heat treatment the samples are held in a sealed chamber under air or argon pressurized to 0.3 MPa. In each case the heat treatment comprised a pre-heating stage and 4 pulses (0.1, 0.25, 1, and 0.5 s), each resulting in a temperature increase to around 3000 K at the surface. During the preheating stage the temperature of the sample was increased to around 1500 K to minimize thermal shock of higher power pulses. After each pulse the sample was left to cool down by dissipating the heat (laser was turned off). Phase transitions were discussed based on the analysis of thermal arrests recorded during the cooling stages. The measurement conditions can be described as quasicontainerless, as only a small spot on the sample surface was laser-irradiated. This effectively prevents contamination by any foreign material. Thus only interactions with the surrounding gas or segregation phenomena can cause local changes in composition.
3. Results Details on all experiments carried out in the frame of this study were listed in Table 1. Example thermograms for MOX14 and MOX62 samples were shown in Fig. 1, where the arrests/inflection points are clearly visible. For each sample two heat treatments were performed (1st and 2nd series) each comprising four shots. The second series were performed on an untreated part of the sample surface. In Fig. 2 the observed thermal arrests were plotted in a T vs Pu content diagram. These values were compared with data reported in ref. [3] and a Calphad prediction of the phase boundaries (liq-
Table 1 Key parameters for all the studied samples. Sample
Used/ discarded thermograms
Lowest recorded arrest (K)
Highest recorded arrest (K)
DT between Tm for stoichiometric MOX and highest arrest
MOX14 MOX24 MOX35 MOX46 MOX54 MOX62
8/0 8/0 8/0 8/0 5/3 8/0
2470 2765 2780 2750 2910 2905
2950 2950 2910 2955 2930 2920
115 75 100 15 65 0
uidus and solidus) for UO1.98 PuO1.96 MOX (calculated using the FuelBase database 2013 version). Eight shots were done for each sample in two series and the results appear to be strongly dispersed for samples weakly doped with Pu. As the Pu content increases, the thermal arrests become less and less scattered. The highest recorded values though are quite similar independent of the Pu composition. In all experiments a dense plume of gases was observed during laser heating. Significant quantities of material evaporated during the increase of temperature and simultaneous oxidation. The vaporization decreased for samples with stronger Pu doping. In some cases exfoliation was observed during of just after the thermal cycles.
4. Discussion Based on the described results we can confidently state that the phase changes had a significant impact on the recorded thermograms. This observation is supported by the fact that in the case of U-rich samples where strong oxidation could be expected, the thermal arrests were dispersed. Also in those cases the thermograms displayed peculiar features stemming from the damaging effects of thermal shocks and oxidation (cracking, exfoliation, pulverization, vaporization). The released plume of gases occasionally partially blocked the emission from the sample’s surface. This happened in the first thermograms reported in Fig. 1 for sample MOX14, where the vapour plume effect can explain the deep apparent temperature decrease preceding the freezing arrest (as demonstrated in dedicated test not reported in this paper). In samples richer in UO2 the temperature of the thermal arrest in air increased with subsequent shots. One explanation might be that the released plume became less dense as the sample’s composition stabilized. Accordingly, the vapour plume-related pre-freezing deep temperature decrease disappeared in the last pulses. It has been shown [4] that even at temperatures as low as 1500 K and for a Pu content of 62 mol%, the products of oxidation in air (patm) can include M3O8. The quantity of this phase in the case of a sample containing 14 mol% Pu was reported to be up to 4 times larger than for 62 mol%. Compared to cubic MO2, this structure is about 40% larger in terms of cell volume [9]. This difference leads to the formation of what is often referred to as the popcorn morphology [10]. The name originates from the peculiar shape of M3O8 precipitates. In a dense pellet or pellet fragment, the formation of sufficient quantities of M3O8 leads to the disintegration and abundant vaporization of the sample surface. In the case of the described experimental approach, the gas in the sample chamber was air pressurized to 0.3 MPa, which implies strongly oxidizing conditions in a part of the considered temperature range. Thus, one might expect oxidation and the formation of even more volatile phases such as UO3 (MO3) apart from M3O8. This would explain the particular features in the thermograms and the dense gas plume in the case of U-rich samples. During subsequent shots, the composition in the affected volume of the sample should stabilize, which in turn would cause a decrease in the vaporization. To verify this hypothesis a high frame rate recording of the sample’s surface during heat treatment would be necessary. This would enable assigning the gas release periods to different phases of the experiment (pre-heating, pulses, or natural cooling). What is also worth noting is the fact that the highest temperatures at which the thermal arrests were observed are quite similar in all cases. Still, these values are much lower than those reported for stoichiometric MOX (see Fig. 2) for U-rich compositions. For Purich compounds, the value approaches the Calphad prediction of the liquidus line and the trend for almost stoichiometric samples [11]. It is true that with the Calphad approach we can only predict
M. Strach et al. / Nuclear Instruments and Methods in Physics Research B 374 (2016) 125–128
127
Fig. 1. Examples of recorded thermograms for MOX14 and MOX62. Each experiments corresponds to 4 subsequent laser pulses (heating–cooling cycles) under air.
Fig. 2. The thermal arrests for all six MOX samples and two series (Air and Ar) compared to results of solidus and liquidus temperatures reported in ref. [3] and a Calphad prediction of solidus and liquidus lines for UO1,98 PuO1,96. The thermal arrests in Ar correspond to those found in the literature. A significant decrease can be observed for melting in air.
equilibria, hence it might seem unjustified to compare the predictions with laser melting results. But, we assume, based on theoretical analysis discussed in a cited article, that the thermal arrests are
representative of some equilibria, as the mobility of species is very high in the considered temperature domain and phase transformations occur quickly. That is, the sample crosses the liquidus or
128
M. Strach et al. / Nuclear Instruments and Methods in Physics Research B 374 (2016) 125–128
solidus surface at a point, to which we can assign a distinct composition, corresponding, within the experimental uncertainty, to the initial one, and place it on the phase diagram. This is supported by the fact that the melting/solidification behavior in Ar is independent of the laser pulse duration and sequence, which should not be the case if we were out of equilibrium. During experiments under air, predicting at which point the sample might cross the solidus and liquidus surfaces is problematic and depends on the sample composition and thermal history. In these cases, the O/M of the samples changed between shots, but the departure from the initial O/M = 2.00 was strongest in the case of U-rich samples. Thus, we should expect the difference between the melting temperature measured for a stoichiometric (O/M = 2.00) under Ar and for a hyper-stoichiometric MOX (O/M > 2.00) under Air to decrease with increasing Pu content, which is the case for our results.
The authors wish to stress the necessity of researching the influence of the O/M on the melting temperature of MOX by developing appropriate experimental approaches. We envisage an extensive experimental campaign to determine quantitatively the effects of melting in air on the structure and cation oxidation state in MOX compounds. We also plan to further use the Calphad approach to estimate the O/M evolution of the sample during the used thermal treatment. Acknowledgements The series of laser melting experiments carried out at the Institute of Transuranium Elements were co-financed by the TALISMAN (Transnational Access to Large Infrastructure for a Safe Management of Actinide) project. References
5. Conclusions Based on laser melting experiments on six MOX samples with Pu content ranging between 14 and 62 mol% in air, we concluded that the phase transformations that take place during the oxidation of the material via species exchange with the atmosphere have a significant impact on the measurement of the temperature at the surface. The phenomena associated with these effects are evaporation of oxygen rich species (UO3 or MO3) and sample damage (due to the formation of M3O8). Although the experimental results were strongly affected by the oxidation of samples, we also conclude that the melting temperature of MOX compounds decreases with an increase in O/M. In the case of MOX14, the decrease can be estimated to be of the order of 100 K.
[1] F. De Bruycker, K. Boboridis, D. Manara, P. Pöml, M. Rini, R.J.M. Konings, Mater. Today 13 (2010) 52. [2] J.J. Carbajo, G.L. Yoder, S.G. Popov, V.K. Ivanov, J. Nucl. Mater. 299 (2001) 181. [3] R. Böhler, M.J. Welland, D. Prieur, P. Cakir, T. Vitova, T. Pruessmann, I. Pidchenko, C. Hennig, C. Guéneau, R.J.M. Konings, D. Manara, J. Nucl. Mater. 448 (2014) 330. [4] M. Strach, R.C. Belin, J.-C. Richaud, J. Rogez, Inorg. Chem. 53 (2014) 12757. [5] T. Truphémus, R.C. Belin, J.-C. Richaud, M. Reynaud, M.-A. Martinez, I. Félines, A. Arredondo, A. Miard, T. Dubois, F. Adenot, J. Rogez, J. Nucl. Mater. 432 (2013) 378. [6] D. Manara, M. Sheindlin, W. Heinz, C. Ronchi, Rev. Sci. Instrum. 79 (2008) 113901. [7] D. Manara, C. Ronchi, M. Sheindlin, M. Lewis, M. Brykin, J. Nucl. Mater. 342 (2005) 148. [8] F. De Bruycker, K. Boboridis, R.J.M. Konings, M. Rini, R. Eloirdi, C. Guéneau, N. Dupin, D. Manara, J. Nucl. Mater. 419 (2011) 186. [9] P. Taylor, D.D. Wood, A.M. Duclos, D.G. Owen, J. Nucl. Mater. 168 (1989) 70. [10] R.J. McEachern, P. Taylor, J. Nucl. Mater. 254 (1998) 87. [11] C. Guéneau, N. Dupin, B. Sundman, C. Martial, J.-C. Dumas, S. Gossé, S. Chatain, F.D. Bruycker, D. Manara, R.J.M. Konings, J. Nucl. Mater. 419 (2011) 145.
Contents lists available at ScienceDirect
Nuclear Instruments and Methods in Physics Research B journal homepage: www.elsevier.com/locate/nimb
Melting behavior of mixed U–Pu oxides under oxidizing conditions Michal Strach a,c, Dario Manara b,⇑, Renaud C. Belin a, Jacques Rogez c a
CEA, DEN, DTEC, SECA, LCC, Cadarache F-13108, Saint-Paul-Lez-Durance, France European Commission, Joint Research Centre, Institute for Transuranium Elements, P.O. Box 2340, 76125 Karlsruhe, Germany c IM2NP, UMR CNRS 7334 – Aix Marseille University, Case 251, Avenue Escadrille Normandie Niemen, 13397 Marseille Cedex 20, France b
a r t i c l e
i n f o
Article history: Received 29 May 2015 Received in revised form 16 December 2015 Accepted 4 January 2016 Available online 10 February 2016 Keywords: MOX U–Pu–O Laser melting Oxidation
a b s t r a c t In order to use mixed U–Pu oxide ceramics in present and future nuclear reactors, their physical and chemical properties need to be well determined. The behavior of stoichiometric (U,Pu)O2 compounds is relatively well understood, but the effects of oxygen stoichiometry on the fuel performance and stability are often still obscure. In the present work, a series of laser melting experiments were carried out to determine the impact of an oxidizing atmosphere, and in consequence the departure from a stoichiometric composition on the melting behavior of six mixed uranium plutonium oxides with Pu content ranging from 14 to 62 wt%. The starting materials were disks cut from sintered stoichiometric pellets. For each composition we have performed two laser melting experiments in pressurized air, each consisting of four shots of different duration and intensity. During the experiments we recorded the temperature at the surface of the sample with a pyrometer. Phase transitions were qualitatively identified with the help of a reflected blue laser. The observed phase transitions occur at a systematically lower temperature, the lower the Pu content of the studied sample. It is consistent with the fact that uranium dioxide is easily oxidized at elevated temperatures, forming chemical species rich in oxygen, which melt at a lower temperature and are more volatile. To our knowledge this campaign is a first attempt to quantitatively determine the effect of O/M on the melting temperature of MOX. Ó 2016 Elsevier B.V. All rights reserved.
1. Introduction Recently, numerous studies have been devoted to the reassessment of the melting temperatures of PuO2, UO2 and mixtures of the two compounds (MOX). Main focus has been put on the minimization of the sample-crucible interactions effects [1], which apparently had affected many previous results [2]. It has been concluded, that in the case of actinides (An), the higher the oxygen potential of a given An–O compound, the higher the discrepancies one might expect between results from experiments using some kind of container and container-less ones [3]. Thus, for UO2 the melting temperatures found in older studies and the new data determined using container-less techniques differ only slightly, while for PuO2 (which has a much higher oxygen potential than the aforementioned compound) the difference was found to be up to 300 K. The melting behavior of MOX with different U/Pu ratios has been thoroughly analyzed by Böhler et al. [3], and it was found that a minimum of the melting temperature curve exists between x(PuO2) = 0.4 and x(PuO2) = 0.7. What was also concluded in this study was the apparent effect of the oxygen stoichiometry ⇑ Corresponding author. http://dx.doi.org/10.1016/j.nimb.2016.01.032 0168-583X/Ó 2016 Elsevier B.V. All rights reserved.
(oxygen to metal ratio O/M) on the results. The described laser heating experiments were carried out either under pressurized argon or air, depending on the predominant element, U and Pu respectively. Although the findings indicated that the impact of O/M variations during thermal treatment were slight and did not have a strong impact on the melting behavior in the presented cases, the authors concluded that further investigations are needed to determine the uncertainties concerning the O/M determination. The cited reference provides compelling argumentation in favor of the accuracy of the presented results concerning stoichiometric compounds (O/M = 2.00 at the beginning of the thermal treatment and at the melting point). But one can imagine a scenario, where the fuel experiences strongly oxidizing or reducing atmospheres or environments before or during melting. In this instance, the melting behavior would surely be affected. In another study [4], we found that oxidation of MOX with Pu contents even as high as 62 mol% results in the formation of significant amounts of oxygen-rich phases, such as MO2+x and M3O8, and possible U/Pu differences in the two resulting phases. This is due to the fact that uranium can increase its valence up to +6 and Pu can be incorporated in the resulting structures.
126
M. Strach et al. / Nuclear Instruments and Methods in Physics Research B 374 (2016) 125–128
In the present work we describe a series of experiments aimed at studying the effect of a strong departure from and O/M = 2.00 insitu on the melting behavior of the studied compounds, and on the applied measurement method itself.
2. Materials and methods Six MOX samples were prepared by co-milling, pressing and sintering stoichiometric UO2 and PuO2 powders in appropriate proportions. The fabrication process was detailed in ref. [5]. Disks were cut form the resulting dense pellets, X mm in diameter and 1–2 mm in height. Stoichiometry was achieved by atmosphere control during sintering (Ar + 5%H2 + H2O) and verified by powder XRD measurements. The Pu content was verified by titration and estimated at 14, 24, 35, 46, 54 and 62 mol%. Details concerning the laser heating setup have been listed in ref. [1,6–8]. Surface temperatures of the studied samples were measured by sub-millisecond resolution pyrometry. The uncertainty was estimated at ±50 K. The equipment used for heating the samples beyond the melting point was a TRUMPFÒ Nd:YAG cw laser radiating at 1064.5 nm, which power–time profiles can be programmed with 1 ms resolution. Pulses ranged in duration from 100 to 1000 ms and power from 180 to 675 W. During heat treatment the samples are held in a sealed chamber under air or argon pressurized to 0.3 MPa. In each case the heat treatment comprised a pre-heating stage and 4 pulses (0.1, 0.25, 1, and 0.5 s), each resulting in a temperature increase to around 3000 K at the surface. During the preheating stage the temperature of the sample was increased to around 1500 K to minimize thermal shock of higher power pulses. After each pulse the sample was left to cool down by dissipating the heat (laser was turned off). Phase transitions were discussed based on the analysis of thermal arrests recorded during the cooling stages. The measurement conditions can be described as quasicontainerless, as only a small spot on the sample surface was laser-irradiated. This effectively prevents contamination by any foreign material. Thus only interactions with the surrounding gas or segregation phenomena can cause local changes in composition.
3. Results Details on all experiments carried out in the frame of this study were listed in Table 1. Example thermograms for MOX14 and MOX62 samples were shown in Fig. 1, where the arrests/inflection points are clearly visible. For each sample two heat treatments were performed (1st and 2nd series) each comprising four shots. The second series were performed on an untreated part of the sample surface. In Fig. 2 the observed thermal arrests were plotted in a T vs Pu content diagram. These values were compared with data reported in ref. [3] and a Calphad prediction of the phase boundaries (liq-
Table 1 Key parameters for all the studied samples. Sample
Used/ discarded thermograms
Lowest recorded arrest (K)
Highest recorded arrest (K)
DT between Tm for stoichiometric MOX and highest arrest
MOX14 MOX24 MOX35 MOX46 MOX54 MOX62
8/0 8/0 8/0 8/0 5/3 8/0
2470 2765 2780 2750 2910 2905
2950 2950 2910 2955 2930 2920
115 75 100 15 65 0
uidus and solidus) for UO1.98 PuO1.96 MOX (calculated using the FuelBase database 2013 version). Eight shots were done for each sample in two series and the results appear to be strongly dispersed for samples weakly doped with Pu. As the Pu content increases, the thermal arrests become less and less scattered. The highest recorded values though are quite similar independent of the Pu composition. In all experiments a dense plume of gases was observed during laser heating. Significant quantities of material evaporated during the increase of temperature and simultaneous oxidation. The vaporization decreased for samples with stronger Pu doping. In some cases exfoliation was observed during of just after the thermal cycles.
4. Discussion Based on the described results we can confidently state that the phase changes had a significant impact on the recorded thermograms. This observation is supported by the fact that in the case of U-rich samples where strong oxidation could be expected, the thermal arrests were dispersed. Also in those cases the thermograms displayed peculiar features stemming from the damaging effects of thermal shocks and oxidation (cracking, exfoliation, pulverization, vaporization). The released plume of gases occasionally partially blocked the emission from the sample’s surface. This happened in the first thermograms reported in Fig. 1 for sample MOX14, where the vapour plume effect can explain the deep apparent temperature decrease preceding the freezing arrest (as demonstrated in dedicated test not reported in this paper). In samples richer in UO2 the temperature of the thermal arrest in air increased with subsequent shots. One explanation might be that the released plume became less dense as the sample’s composition stabilized. Accordingly, the vapour plume-related pre-freezing deep temperature decrease disappeared in the last pulses. It has been shown [4] that even at temperatures as low as 1500 K and for a Pu content of 62 mol%, the products of oxidation in air (patm) can include M3O8. The quantity of this phase in the case of a sample containing 14 mol% Pu was reported to be up to 4 times larger than for 62 mol%. Compared to cubic MO2, this structure is about 40% larger in terms of cell volume [9]. This difference leads to the formation of what is often referred to as the popcorn morphology [10]. The name originates from the peculiar shape of M3O8 precipitates. In a dense pellet or pellet fragment, the formation of sufficient quantities of M3O8 leads to the disintegration and abundant vaporization of the sample surface. In the case of the described experimental approach, the gas in the sample chamber was air pressurized to 0.3 MPa, which implies strongly oxidizing conditions in a part of the considered temperature range. Thus, one might expect oxidation and the formation of even more volatile phases such as UO3 (MO3) apart from M3O8. This would explain the particular features in the thermograms and the dense gas plume in the case of U-rich samples. During subsequent shots, the composition in the affected volume of the sample should stabilize, which in turn would cause a decrease in the vaporization. To verify this hypothesis a high frame rate recording of the sample’s surface during heat treatment would be necessary. This would enable assigning the gas release periods to different phases of the experiment (pre-heating, pulses, or natural cooling). What is also worth noting is the fact that the highest temperatures at which the thermal arrests were observed are quite similar in all cases. Still, these values are much lower than those reported for stoichiometric MOX (see Fig. 2) for U-rich compositions. For Purich compounds, the value approaches the Calphad prediction of the liquidus line and the trend for almost stoichiometric samples [11]. It is true that with the Calphad approach we can only predict
M. Strach et al. / Nuclear Instruments and Methods in Physics Research B 374 (2016) 125–128
127
Fig. 1. Examples of recorded thermograms for MOX14 and MOX62. Each experiments corresponds to 4 subsequent laser pulses (heating–cooling cycles) under air.
Fig. 2. The thermal arrests for all six MOX samples and two series (Air and Ar) compared to results of solidus and liquidus temperatures reported in ref. [3] and a Calphad prediction of solidus and liquidus lines for UO1,98 PuO1,96. The thermal arrests in Ar correspond to those found in the literature. A significant decrease can be observed for melting in air.
equilibria, hence it might seem unjustified to compare the predictions with laser melting results. But, we assume, based on theoretical analysis discussed in a cited article, that the thermal arrests are
representative of some equilibria, as the mobility of species is very high in the considered temperature domain and phase transformations occur quickly. That is, the sample crosses the liquidus or
128
M. Strach et al. / Nuclear Instruments and Methods in Physics Research B 374 (2016) 125–128
solidus surface at a point, to which we can assign a distinct composition, corresponding, within the experimental uncertainty, to the initial one, and place it on the phase diagram. This is supported by the fact that the melting/solidification behavior in Ar is independent of the laser pulse duration and sequence, which should not be the case if we were out of equilibrium. During experiments under air, predicting at which point the sample might cross the solidus and liquidus surfaces is problematic and depends on the sample composition and thermal history. In these cases, the O/M of the samples changed between shots, but the departure from the initial O/M = 2.00 was strongest in the case of U-rich samples. Thus, we should expect the difference between the melting temperature measured for a stoichiometric (O/M = 2.00) under Ar and for a hyper-stoichiometric MOX (O/M > 2.00) under Air to decrease with increasing Pu content, which is the case for our results.
The authors wish to stress the necessity of researching the influence of the O/M on the melting temperature of MOX by developing appropriate experimental approaches. We envisage an extensive experimental campaign to determine quantitatively the effects of melting in air on the structure and cation oxidation state in MOX compounds. We also plan to further use the Calphad approach to estimate the O/M evolution of the sample during the used thermal treatment. Acknowledgements The series of laser melting experiments carried out at the Institute of Transuranium Elements were co-financed by the TALISMAN (Transnational Access to Large Infrastructure for a Safe Management of Actinide) project. References
5. Conclusions Based on laser melting experiments on six MOX samples with Pu content ranging between 14 and 62 mol% in air, we concluded that the phase transformations that take place during the oxidation of the material via species exchange with the atmosphere have a significant impact on the measurement of the temperature at the surface. The phenomena associated with these effects are evaporation of oxygen rich species (UO3 or MO3) and sample damage (due to the formation of M3O8). Although the experimental results were strongly affected by the oxidation of samples, we also conclude that the melting temperature of MOX compounds decreases with an increase in O/M. In the case of MOX14, the decrease can be estimated to be of the order of 100 K.
[1] F. De Bruycker, K. Boboridis, D. Manara, P. Pöml, M. Rini, R.J.M. Konings, Mater. Today 13 (2010) 52. [2] J.J. Carbajo, G.L. Yoder, S.G. Popov, V.K. Ivanov, J. Nucl. Mater. 299 (2001) 181. [3] R. Böhler, M.J. Welland, D. Prieur, P. Cakir, T. Vitova, T. Pruessmann, I. Pidchenko, C. Hennig, C. Guéneau, R.J.M. Konings, D. Manara, J. Nucl. Mater. 448 (2014) 330. [4] M. Strach, R.C. Belin, J.-C. Richaud, J. Rogez, Inorg. Chem. 53 (2014) 12757. [5] T. Truphémus, R.C. Belin, J.-C. Richaud, M. Reynaud, M.-A. Martinez, I. Félines, A. Arredondo, A. Miard, T. Dubois, F. Adenot, J. Rogez, J. Nucl. Mater. 432 (2013) 378. [6] D. Manara, M. Sheindlin, W. Heinz, C. Ronchi, Rev. Sci. Instrum. 79 (2008) 113901. [7] D. Manara, C. Ronchi, M. Sheindlin, M. Lewis, M. Brykin, J. Nucl. Mater. 342 (2005) 148. [8] F. De Bruycker, K. Boboridis, R.J.M. Konings, M. Rini, R. Eloirdi, C. Guéneau, N. Dupin, D. Manara, J. Nucl. Mater. 419 (2011) 186. [9] P. Taylor, D.D. Wood, A.M. Duclos, D.G. Owen, J. Nucl. Mater. 168 (1989) 70. [10] R.J. McEachern, P. Taylor, J. Nucl. Mater. 254 (1998) 87. [11] C. Guéneau, N. Dupin, B. Sundman, C. Martial, J.-C. Dumas, S. Gossé, S. Chatain, F.D. Bruycker, D. Manara, R.J.M. Konings, J. Nucl. Mater. 419 (2011) 145.