Community Research and Development Information Service - CORDIS

FP7

ASGARD Report Summary

Project ID: 295825
Funded under: FP7-EURATOM-FISSION
Country: Sweden

Final Report Summary - ASGARD (Advanced fuelS for Generation IV reActors: Reprocessing and Dissolution)

Executive Summary:
Now at the end of the project it is clear that ASGARD was a success on many levels. The aims were originally multi-dimensional with the overall focus on the recyclability of novel nuclear fuels. The main dimensions were: the scientific achievements, investigating how to increase the industrial applicability of the fabrication of these novel fuels, the bridging of the often separate physics and chemical communities when dealing with nuclear fuel cycles and last but not least to offer an extensive education and training to younger scientists including a broadening of their experience by visiting and performing work at other facilities than their own. At this point it is very satisfying to note that all these objectives were met with great success.
The scientific work was divided into domain representing the different fuel types investigated, i.e. oxides including inert matrix fuels, nitride fuels and carbide fuels. During the course of the project there were a large number of scientific achievements evidenced by the number of scientific publications in peer-reviewed journals, 27. It needs, however to be noted that this is only the number of published paper at the end of the project. The real number is estimated to be about the double since many achievements were reached rather late in the project.
The success of the integration and cross disciplinary training is shown by the successful implementation of the Travel Fund. It allowed that 4 young scientists were each spending more than 1 month at another partner laboratory, funded 23 young scientists to visit conferences and meetings presenting their achievements as well as the organisation of 3 dedicated summer/winter schools having about 15 participants each which in the case of the last one dealing with plutonium chemistry was the absolute maximum possible. Therefore a follow up of this school is asked for internationally since there is essentially only one university laboratory in Europe that can handle such a course.
For the scientific domains there are naturally many different achievements but among the more important ones by domain are as follows.
In the oxide domain the recyclability of the inert matrix fuels based on MgO or Mo was investigated and it was clear that it is possible to both dissolve the fuels in a good way but also to pre-treat the dissolution liquor in a way that it can be fed into the already existing reprocessing process. Moreover, it was shown that fabrication of pure 50/50 (Pu,Am)O2 was possible. However, its dissolution properties was not investigated but suggested for a follow up project.
In the nitride domain it was shown that the pivotal cost for production of 15N can be reduced significantly but is still a bit high for competitive fuel production. UN production technique was optimised to obtain a total control of e.g. oxygen impurities and controlling porosity obtaining a 99.88% dense pellet. A significant achievement was also the production of pure PuN by using both nitriding from oxides and by sol-gel rote. It is unsure if this kind of material was ever produced by these means before.
In the carbide domain the alleged pyrophoricity of carbide material was investigated showing that by controlling oxygen and water access even fine power can be handled safely. The parameters for ignition of both power and pellets were determined thus increasing the safety of this kind of fuel handling significantly. A new sol-gel technique based on microwave gelation was also developed. By this ground breaking invention the industrial applicability of the sol-gel technique for production of any fuel precursor from solution was increased significantly. Now there is no need for washing, drying etc the microspheres and also the production of contaminated oils is prevented.

Project Context and Objectives:
The Strategic Energy Technology plan (SET-plan) of the European Commission identifies fission energy as an important contributor to meet long-term objectives for reduction of greenhouse gas emissions. Sustainability of nuclear power may be achieved by the introduction of fast neutron Generation IV reactors and their associated fuel cycle facilities, as described in the Strategic Research Agenda (SRA) of the Sustainable Nuclear Energy Technology Platform (SNE-TP).
The ESNII task force of SNE-TP (European Sustainable Nuclear Industrial Initiative) has defined a road-map towards the demonstration of sustainable Generation IV systems. This plan foresees the construction and operation of one prototype sodium cooled reactor (ASTRID) with a power of 600 MWe, two demonstration reactors using lead and gas coolant, respectively (ALFRED and ALLEGRO), a lead-bismuth cooled materials test and irradiation facility (MYRRHA), a minor actinide capable fuel fabrication pilot plant (ALFA) and other supporting facilities.
Generation IV reactors are foreseen to have a breeding ratio of plutonium equal to unity, while functioning as burners of minor actinides. The use of novel coolant technologies and advanced fuels in combination with stringent safety objectives of Generation IV systems requires significant R&D to be carried out in the immediate future, in order for demonstration on industrial scale to become possible in the next decade. Relevant research is performed in national programmes as well as in FP7 projects such as ESFR, LEADER, GOFASTR, ACSEPT, GETMAT, FAIRFUELS, FREYA and F-BRIDGE. Unfortunately, today, integration between reactor, fuel and recycling communities is lacking, in some cases resulting in discrepancies between the reactor design on one hand, and the technological feasibility of fabricating, dissolving and reprocessing the selected fuel on the other hand. This is reasonably true for MOX fuel, but even more evident for advanced nuclear fuels. Such future nuclear fuels comprise e.g. inert matrix fuels, nitride fuels and carbide fuels. In all these areas, there are still large gaps in knowledge before any process for the manufacturing, operation and recycling of these fuels can take place.

Consistently with the above mentioned future nuclear research, the ASGARD project's main objective is to provide a structured R&D framework bridging the research on fuel fabrication and reprocessing issues. The main focus will lie on future fuels for a sustainable nuclear fuels cycle. The main problem today is to tie the recycling of the nuclear fuel to the fabrication of new fuels. Seen in this context the outline of the work on each of the fuel types will be: Dissolution (of irradiated and unirradiated fuel), Conversion and Fabrication.

Figure 1 The sustainability circle for nuclear fuel where ASGARD fills the gap between the main focus of FP 7 ACSEPT and FP7 FAIRFUELS

These processes will be applied to the different fuel types that have been identified as possible future alternatives for the next generation of power producing reactors:
1. Oxide and CerCer / CerMet Inert Matrix Fuels
2. Nitride Fuels
3. Carbide Fuels

Oxides
The oxide dissolution and separation strategy is a fairly mature process being dealt with and optimised in the FP7 EURATOM ACSEPT project. New separation strategies have been tested on genuine spent fuel and the selected processes will be evaluated for industrial implementation. Whereas the above is valid for actinide oxide fuels, such as MOX and/or Minor Actinide containing MOX, the dissolution and separation issues for inert matrix fuels containing ceramic MgO or metallic molybdenum (Mo), has not been investigated coherently. In order to be complementary to ACSEPT, the ASGARD project focuses therefore mainly on the Inert Matrix Fuels (IMF) with molybdenum or magnesium-oxide. It is of crucial importance to take into account the behaviour of the matrix elements in the dissolution and separation processes and to check their compatibility with the future immobilisation (impact on the stability of the waste and amount of generated waste).
In addition, the use of molybdenum as an inert matrix poses additional challenges with respect to its redox chemistry, the need to avoid precipitation or co-precipitation, and the necessity to recover the isotopically tailored Mo material.
For the assessment of inert matrix fuels, some basic studies will be performed to assess the dissolution of minor-actinide containing oxides, specifically with high americium and plutonium content.
In this Domain, the last objective to be addressed is the conversion of the reprocessed solution to suitable precursors for fuel fabrication. Up to now, these conversion issues have not been investigated systematically. They form a crucial connecting part in the sustainability circle for nuclear fuel, see Figure 1.

Nitrides
Nitride fuels constitute a high performance alternative to oxide fuel. Major advantages include a higher actinide density and a combination of high thermal conductivity with high melting temperature. The latter are particularly important in the context of transmutation in Generation IV reactors, since the addition of minor actinides to the fuel is detrimental for reactivity feedbacks. The resulting increase in fuel temperature under power transients is more easily accommodated by the larger margin to failure of the nitride fuel. The solubility of actinide nitrides in nitric acid is good in general. A major issue that needs to be addressed is how 15N recycling is to be implemented. Already in the case of oxide fuels, 14C is a major contributor to the dose commitment arising from the operation of reprocessing plants. A large scale use of nitride fuel in a closed fuel cycle would increase the 14C production above regulatory limits if enrichment of nitrogen in 15N is not undertaken. The cost for this enrichment might however become a significant penalty if 15N is not recycled. Applying standard dissolution processes, dilution of 15N in nitric acid is unavoidable. Hence a pre-treatment step involving voloxidation of the nitride fuel and recovery of the 15N stream should be performed. Similarly, losses of 15N at the head end of the fabrication process must be minimized. A possible solution is the use of direct hydridation/nitridation of metallic actinides, which permits a closed gas cycle to be applied.
The ASGARD project will address all these issues, including improved processes for enrichment, the impact of fabrication route on dissolution performance and recovery of 15N.

Carbides
Carbide fuels have been considered as fuels for Fast Reactors since the 1960s and many irradiation experiments on U and U/Pu carbide have been performed globally, with recent successes particularly for the Fast Breeder Test Reactor at Kalpakkam, India. The physical properties of carbide fuel make them attractive because they are conducive to high specific rod powers with relatively low fuel centre temperatures: start of life power capability is increased, power-to-melt margin is increased and fatter (more economic) pins are facilitated. Because fuel temperatures are low, the fuel suffers little or no restructuring and the release of fission gases and volatile fission products is low. Under transient conditions, high thermal conductivity and low thermal expansion offer certain advantages i.e. lower fuel/clad mechanical interaction (FCMI) levels in over power transients and lower cladding temperatures in under cooling events. Operationally, better internal breeding reduces reactivity decline over the fuel lifetime which eases control rod management.
Carbide fuel chemical properties provide the advantage that under normal operating conditions fuel/stainless steel clad chemical interaction is virtually absent; additionally, swelling due to fuel/coolant interaction in a failed pin should not occur to the benefit of minimising contamination of the reactor circuit. Thus two sources of concern present with oxide pins are potentially eliminated. On the down side there are a number of evident problems with carbide fuel. Performance limitations are of concern because of the significant potential for unacceptable fuel/clad mechanical interaction (FCMI) due to the high swelling propensity from gases trapped within the fuel matrix leading to the possibility of extensive reaction between fuel and cladding at higher transient temperatures, thus limiting the carbide fuel centreline pellet temperature and the overall economic performance of the reactor. It is then clear that if the fuel swelling mechanism can be avoided/reduced carbide fuel could have significant advantages towards FCMI compared to oxide fuel.
In the fuel cycle, manufacturing involves the handling of pyrophoric powders, which may not be amenable to scale-up; fuel reprocessing is problematic since mixed U/Pu carbide dissolution results in undesirable side products in nitric acid.
ASGARD will seek solutions to the problems of fuel swelling and address the issues concerning the reprocessing of carbide fuels. The objectives are to fabricate carbide fuel in such a way that avoids fuel swelling and identify a suitable route for reprocessing of carbide fuels.

Due to the various stages of development for the different fuel types, the ASGARD project will have different goals for the different fuels. However, in principle, the aim is to reach an understanding of the manufacturing and recycling of the different kind of fuels to a similar scientific level as for the oxide fuels. If, during the course of the project, decisions are taken for a selection of a specific fuel, the focus will shift towards the problems associated with that particular fuel. However, the research on the other types of fuel will continue. The broad scope of ASGARD ensures that relevant knowledge is generated regardless of which of the fuel alternatives will be the dominating one in the future. There are clear cross cutting activities in the field of conversion for all three types of the fuel. Once the actinide material is in solution, the conversion to suitable base solids may be very similar regardless of the origin of the dissolved fuel type.

It is evident that the content of ASGARD has close links to FP7 projects ACSEPT and FAIRFUELS. But it should be noted that ASGARD explicitly aims at complementing the R&D that is performed in these projects, and in many cases also at bridging the gap between the two. There should be no overlap or duplication of the activities performed in other projects. This is assured already in the preparation stage, where both coordinators of ACSEPT and FAIRFUELS were actively involved, as well as by maintaining close cooperation with ACSEPT and FAIRFUELS during the course of the project.

Training and education
A training and education programme will complement the main R&D programme, which aims to share the acquired knowledge among communities and generations, and maintain the nuclear expertise at the fore-front of Europe. The training and educational work package of ASGARD will work in close collaboration or will follow up other training programmes (like e.g. CINCH, ENEN and IAEA) for maximum gain and efficiency in the knowledge dissemination and strengthening the European human capital in this area.

Project Results:
1.3 Description of the main S&T results/foregrounds
The ASGARD project has been a clear success with respect to both the scientific and educational achievements during the implementation of the project but also for the ability to keep the deadlines and research structure originally planned. Considering many other projects involving work at highly radioactive facilities and including irradiated material this is a clear and significant achievement in itself. The only deviation was an unexpected closure of the hotlab at NRG for 6 months. This was, however, well handled by the management of the ASGARD project as well as at NRG and thus causing only minor delays.
In the following sections the major achievements throughout the project will be presented domain by domain. However, due to the limited space and the fact that most results are of a novel scientific nature the details of the work are given in dedicated scientific publications to which the interested reader is directed.
It is important to note already at this point that the outcomes of a project like ASGARD is not only the significant scientific findings and development but also the actual building up of competence and keeping relevant facilities active. Today there are very few infra-structures (and essentially only one University) that can handle the type of activities performed in the ASGARD project and a diminishing amount of people with the appropriate knowledge about working safely and securely with significant amount of radioactive material. For these reasons other important achievement of the ASGARD project is the investment in human capital and resources. There is a number of young scientists that though the ASGARD project had the ability to train with real radioactive material to levels which are relevant for real nuclear activities and thus ASGARD was clearly contributing the keeping of know-how and expertise in this field in Europe.

1.3.1 Domain 1: Management, education and training, dissemination

1.3.1.1 Introduction
This domain is divided into 3 work packages. However, since the project management, WP1.1, is reported separately within the 3rd EC Periodic Report it is not included in this report.

1.3.1.2 WP1.2 Education, training, mobility
One of the main challenges facing the nuclear sector today is the loss of competence as the older generation retire. The reason for this is typically that the future of nuclear power has been strongly debated and thus seen as a somewhat unsecure field for the young ones to start in. In addition, work with real amounts of radioactive material is a highly complex procedure both in the context of actual physical handling regarding safety and security but it is also surrounded by a strong legislative framework. The latter leading to the fact that the number of facilities actually able to perform relevant work with radioactive material is very low. In the University world they are even fewer. In Europe today it is essentially only the Chalmers University laboratory that can handle reasonable amounts of both alpha, beta and gamma radioactivity within its premises.
The ASGARD project had the perfect aim for helping this education and training need in Europe since the focus was to actually to manufacture and investigate real nuclear fuel material. One significant advantage was the fact that both chemistry and physics sciences were mixed as well as solid state and solution chemistry. This showed to be the perfect base for interdisciplinary training and education in many aspects of the handling of radioactive material. In addition to this, there was a strong collaboration between ASGARD and other relevant EU project running more or less in parallel such as CINCH, FAIRFUELS, TALISMAN and SACSESS. Thus all students in these projects had the possibility to gain from the training and education offered in the extensive program of ASGARD.
Apart from the normal activities such as scientific meeting there were other education and training activities actively offered and supported by the ASGARD project. These include an extensive mobility program for giving the possibility to make longer stays at other partner laboratories to widen the knowledgebase of the young scientist but ASGARD also offered summer/winter schools as well as training sessions for dedicated areas connected to some of the project meetings. Since these training sessions became considerably more extensive than originally anticipated they will be reported together with the summer/winter schools in this text.

1.3.1.2.1 Mobility
The mobility scheme of ASGARD had two main components: the possibility to make a longer visit in another partner’s laboratory and to visit conferences or external courses to present research material or gain relevant knowledge.
In the case of the laboratory visits this was used by 5 young researchers. Two different persons were going from the ICHTJ in Poland to Forschungszentrum JÜLICH at two different occasions. The main aim was to try the methods developed at ICHTJ on relevant solutions available in JÜLICH. One PhD student was travelling from JÜLICH to KIT for investigating the dissolution behaviour of inert matrix fuels. These fuels were produced in JÜLICH and the actual dissolution was made in KIT. In order to investigate methods developed using only uranium on more relevant material such as plutonium one PhD student travelled from KTH to the Chalmers facilities. This was also the focus of the last PhD student in this program. She was developing photochemical precipitation methods at CTU using mainly thorium and this method was extended to gamma irradiation at Chalmers where also EXAFS studies of the produced material were made. This last project was collaboration between the ASGARD and TALISMAN projects.
The mobility grant for visiting external courses or relevant conferences was used by 24 younger scientists. Thus it is clear that a significant increase in mobility of these future scientists was achieved through the ASGARD scheme.

1.3.1.2.2 Summer/winter schools and training events
There were four main training and education events organised during the ASGARD project. They were spread out through the duration of the project and also diverse in the sense of theme and organising partner.

Fabrication methods & Irradiation performance
The first summer school (named Winterschool 2013) was held on 29 – 30 January 2013 at NRG, Petten, the Netherlands. This course theme was 'Fabrication methods & Irradiation performance' and was organized in cooperation with the 7th FP projects FAIRFUELS (http://www.fp7-FAIRFUELS.eu) and CINCH (http://cinch-project.eu).
The organisation was mainly done by NRG together with Evalion, CTU and Chalmers
The program was as follows:

Tuesday 29 January 2013 – Fabrication methods & Fuel irradiation experiments
9:00 - 9:15 Welcome at NRG
9:15 - 12:30 Industrial fuel manufacturing course – Hans Widegren, Westinghouse
13:30 - 14:30 Sol gel methods & oxide/nitride sphere-pac fabrication– Manuel Pouchon , PSI
14:30 – 15:30 Resin process for making oxide and carbide microspheres - Sébastien Picart, CEA
15:45 - 17:00 Irradiation studies and post-irradiation examinations of advanced fuel materials, Ralph Hania, NRG

Wednesday 30 January 2013 – Final Public Workshop CINCH
9:00 - 11:00 Workshop CINCH – part I
11:15 – 12:30 Workshop CINCH – part II
13:30 - 17:00 Visit to Hot cell laboratories/Decommissioning and waste treatment facility

The leaflet containing all necessary information (program, venue, travel fund) was made to allow the announcement of the First ASGARD School on the official websites of all projects, on the websites of the institues and universities involved in the projects, on the websites of other EU projects and other related websites.

The number of participants to this first ASGARD event was 27 /35 during the first/second day. Each participant obtained a booklet with the hand-outs of the presentations, including short information about the lecturer. A part of the First ASGARD School was a visit of the NRG nuclear facilities – Hot Cells Laboratories and Decommissioning and waste treatment facility.

Fuel fabrication and reprocessing
In order to set the stage for the general aim of the ASGARD project the coupling between the manufacturing/properties of the different advanced nuclear fuels and their reprocessing a short teaching activity was launched in June 2013 in Warsaw. The main aim was with respect to the safety concerns given by specialists in the respective domains. For the purpose of this training session, these lectures were combined with a short “Reprocessing course” organised jointly by NNL and KIT. This internal school was targeted at ASGARD Ph.D. students and young researchers and should have given them an overview of the status quo in the field of nuclear fuels production and some basic knowledge of the reprocessing issues. The number of participants was 18.

The programme was as follows:

• Lecture 1 – OXIDE FUELS (NRG, E.De Visser-Týnová)
• Lecture 2 – NITRIDE FUELS (KTH, M.Jolkkonen)
• Lecture 3 – CARBIDE FUELS (PSI, N. Shcherbina)
• Lecture 4 – Reprocessing Course - Part A (NNL, M. Sarsfield, Ch. Maher)
• Lecture 4 – Reprocessing Course - Part B (KIT-INE, A. Geist)

Fuel characterisation and isotopic separation
The second training event was organised in conjunction with the ordinary 2014 ASGARD progress meeting in the January 2014 in Stockholm. This training session combined two activities:

• Fuel Characterization course given by KTH (1.5 half days including the laboratory exercises)
• Stable Isotope Separation course given by INCDTIM (half day course).
The participation at this course was rather large exceeding 20 students and the diverse topics handled provided a highly appreciated combination of both lecturing and hands on training in the laboratory. The final programme of the training session was as presented in Table 1.

All the lectures of this course were recorded, converted to podcasts, and will be used for the e-learning purposes.

Working with plutonium
The 2nd ASGARD Summer School and thus the fourth education and training event took place at Chalmers University of Technology in Gothenburg, Sweden, from 4th to 8th of May 2015. Its original topic was modified to suit better the demand – the proposed “Sol gel technique and fuel fabrication training” was replaced by “Working with Plutonium”. The School was a joint effort between the SACSESS, ASGARD and CINCH-II FP7 EURATOM projects but mainly organised through the ASGARD community. A similar joint course between ASGARD and SACESS was given by the NNL and at that time mainly organised by the SACSESS community.
The aim of the school has been to give PhD students and young researches the possibility of hands-on training with Pu in a safe and didactic manner. Since there are very few facilities that can offer the possibility for real hands on training working with this kind of material on a safe and secure way the interest in the community was very high. The maximum number of student was originally set to 10 but due to the demand it was increased. There were 12 students from 5 different countries: Italy, UK, Germany, Czech Republic and The Netherlands attending. The teachers were the seniors and PhD students from Chalmers University involved in all or some of the organizing projects. Due to the high demand for this course it is expected that a repetition will be suggested in an upcoming project.
Part of the didactic material used for the school has been provided by CINCH-II project, while other parts were tailored specially for this event. Even more, the lectures were filmed and prepared for future use by CINCH-II staff as e-learning material for future use. They are now available at the CINCH Moodle at http://moodle.cinch-project.eu.

The Summer School comprised:
Lectures:
1. Radiation Protection
2. Plutonium Chemistry
3. Plutonium Analysis and quantification
Hands-on training:
1. Radiation Protection
2. Glove-box training
3. Plutonium oxidation and extraction
4. The internal gelation process
5. Thermal treatment, pellet fabrication and characterization.

All the lectures of this course were recorded, converted to podcasts, and will be used for the e-learning purposes.

1.3.1.3 WP1.3 Dissemination, exploitation and networking
The dissemination of the knowledge generated in a project like ASGARD is divided into several parts. Naturally for a scientific project the main channel for spreading the knowledge is through scientific publications. By the pure nature of the ASGARD project and the fact that it is the first of its kind most of the publications are published rather late. Thus at the writing of this report 17 scientific papers were published in reviewed journals (additional 10 submitted or accepted). It is expected that this number will be about doubled when the total outcome of the project is published. Involvement of young generation of scientists into the research programme of ASGARD resulted also into 8 defended Master or PhD theses during the project (additional ones are to be defended after the project end).
Another important channel for spreading the knowledge generated is though conference presentations. In this context the ASGARD project has been working two fold. Both though the support generated by the mobility plan, as discussed above, but also though active promotion of the younger participants to present their results by themselves at different relevant scientific events. The total number of such conference presentations though the duration of ASGARD is 61 for oral presentations at conferences and 20 papers in proceedings of such events.
In order to increase the visibility of the ASGARD project as well as to gain important external input to the work performed three international events were organised. Originally some of these were planned to be stand-alone events but it became clear that the wealth of scientific meetings and conferences was significant why it was deemed more efficient to join ASGARD session to already existing, high quality conferences. The conferences thus including dedicated ASGARD sessions were: ATALANTE (2012), RADCHEM (2014) and TOPFUELS (2015). In the latter case the session was co-organised by the PELGRIMM project. In these sessions which typically were between 0.5- 1 day the participation between project presentations and other external, relevant presentations was about 50/50. This division ensure that both the objectives of ASGARD visibility as well as important external input were met in a satisfactory manner.
A special link was kept with CINCH-II project as already mentioned in WP1.2 description. ASGARD and CINCH-II organized joint project meetings, exchanged information and used their results in highly synergic manner.
Some of the training events were also organized in cooperation with FAIRFUELS, SACSESS and ACSEPT projects as mentioned above.

The project generated three patent applications:
1. CEA: Method for colloidal preparation of a metal carbide, said metal carbide thus prepared and uses thereof – WO2016020496
2. ICHTJ: Method for preparing of uranium-neodymium dioxide spherical grains Double Extraction Process – PL403817
3. ICHTJ: A method for producing uranium carbide of spherical and irregular grains as a precursor to fuel for reactors of the new, fourth generation – PL414768

Additionally, 31 Exploitable Foregrounds have been identified and described by ASGARD partners and reported through D1.3.7 Exploitation Plan including their plans for further use in various domains (mainly for further research but also for education and training, commercialization and other purposes). Description of the Exploitable Foregrounds is a part of this Final Report.

Dissemination task generated in total more than 150 dissemination actions including apart from the scientific publications mentioned above also 28 poster presentations and numerous other dissemination measures (press releases, generic project presentations, project events, webpage and others) as listed in other part of this Final Report.

The ASGARD project webpage at http://asgardproject.eu attracted 9 130 unique visitors during its operation who generated more than 33 000 page views to regularly updated content of the web. This number is expected to increase significantly with recent publication of (publishable) outputs of the project on the webpage. The webpage will be continued after the project end and the statistics of its visits will be recorded.

1.3.2 Domain 2 Oxide fuels and inert matrix fuels including CERCER and CERMET

1.3.2.1 Introduction
The oxide dissolution and separation strategy is a fairly mature process being dealt with and optimised in the FP7 EURATOM ACSEPT project. New separation strategies have been tested on genuine spent fuel and the selected processes have been evaluated for industrial implementation. Whereas the above is valid for actinide oxide fuels, such as MOX and/or Minor Actinide containing MOX, the dissolution and separation issues for inert matrix fuels containing ceramic MgO or metallic molybdenum (Mo), has not been investigated coherently. In order to be complementary to ACSEPT/SACSESS the ASGARD project focused therefore mainly on the Inert Matrix Fuels (IMF) with molybdenum or magnesium-oxide. It is of crucial importance to take into account the behaviour of the matrix elements in the dissolution and separation processes and to check their compatibility with the future immobilisation (impact on the stability of the waste and amount of generated waste).
In addition, the use of molybdenum as an inert matrix poses additional challenges
with respect to its redox chemistry, the need to avoid precipitation or co-precipitation,
and the necessity to recover the isotopically tailored Mo material.
For the assessment of inert matrix fuels, some basic studies have been performed to assess the dissolution of minor-actinide containing oxides, specifically with high americium and plutonium content.
In this Domain, the last objective to be addressed is the conversion of the reprocessed solution to suitable precursors for fuel fabrication. Up to now, these conversion issues have not been investigated systematically. They form a crucial connecting part in the sustainability circle for nuclear fuel.

1.3.2.2 WP2.1 Inert matrix fuels
The dissolution and separation issues for inert matrix fuels containing metallic molybdenum (CerMet) has been investigated with focus on the Inert Matrix Fuels (IMF) with molybdenum oxide where, except for the manufacturing, the handling of the inert component could be of a major concern in a recycling process.
Fresh fuels with CeO2, UO2 and PuO2 in the molybdenum matrix have been prepared and characterized. These fuels have been dissolved close to the PUREX process conditions, in some cases with addition of Fe (III) to speed up the dissolution. The dissolution rate of Mo was strongly dependent on the acid concentration as well as the temperature, as seen in Figure 4.

The dissolution of Mo pellets in HNO3 was complex due to its specific aqueous redox chemistry (Mo -II - +VI) and went along with precipitation phenomena due to Mo polymerization at high acid and Mo concentrations. The addition of 1 mol/L ferric nitrate to the acid can prevent precipitation at low nitric acid concentrations and accelerate the dissolution. However, it is unclear whether adding such an amount of Fe(III) to the dissolver solution could influence the subsequent extraction process. The trends of the dissolution kinetics of molybdenum in nitric acid without Fe(III) or containing 1 equivalent of Fe(III) per equivalent of Mo could be confirmed with mixed Mo/CeO2 (60/40 wt.%) pellets at room temperature. The dissolution of cerium dioxide is not influenced by Fe(III). Semi-warm dissolution tests revealed that while the solubility of molybdenum in nitric acid increases in the presence of iron, the solubility of the actinides is decreased. In all cases less than 5% of the actinides dissolved. A dissolution study on irradiated fuels (Pu0.8Am0.2)O2 in Mo matrix has been performed as well, in both cases of Pu fuels it seems that the matrix can be dissolved; remaining residue is expected to be PuO2.
The dissolution behaviour of fuel (Pu0.8Am0.2)O2 in Mo matrix irradiated at HFR Petten has been studied. The fuel pellets have been cut in slices of 3 mm and dissolved, four dissolution experiments have been performed. The dissolution could be divided in three steps:
1) dissolution of the molybdenum matrix; 2) isolation of active material and 3) dissolution of actinides oxides. The first step of the dissolution – removal of the Mo matrix – was developed to avoid any precipitation of molybdenum at a later stage. The Mo matrix was dissolved in 4 mol/L HNO3 at ambient temperature. The dissolved material was removed. Dissolution of the actinide oxide material in boiling 8 mol/L HNO3 with addition of HF or Ag(II) has been performed. None of the tests has led to a complete dissolution of the material; a black residue remained as in the case of the fresh fuel dissolutions.
Dissolution of the actinide oxide material in boiling 8 mol/L HNO3 with addition of HF or Ag(II) has been performed. None of the tests has led to a complete dissolution of the material; a black residue remained as in the case of the fresh fuel dissolutions.
The separation of the matrix from the fuel by thermal treatment was considered as an alternative to the very complex dissolution of Mo based IMF. Molybdenum was quantitatively recovered as pale yellow crystals, which were identified as MoO3 by XRD. In the case of thermal treatment of Mo/PuO2 mixtures at equivalent conditions a mixed oxide PuMo2O8 is expected to form, which will hinder the separation. However, PuMo2O8 is well soluble in nitric acid. The thermal treatment of the fuel is a promising method for the separation of the bulk molybdenum to simplify the dissolution and minimize the effect on the liquid-liquid extraction.
A direct measurement of the Mo species in aqueous solutions at high acidic strengths was achieved by means of nano ESI-MS. The dissolution kinetics of Mo show, that low concentrations and low timespans leads to rather simple species distributions. With further dissolution of Mo, larger oligomers are formed with up to 10 Mo-moieties. When the Mo–pellet completely dissolved, the species distribution is rather complicated. On the other hand, at lower acidic strength samples the tendency towards formation of larger polymeric species increases. Each of the many different species can exhibit a different behaviour during liquid-liquid extraction steps. The addition of ferric nitrate to the acid can increase the rate of dissolution of Mo pellets. From the results of the Mo and Fe samples, we can deduce that mixed Mo and iron species are formed in the solution. These might be responsible for the increased dissolution rate. In mixed Mo and Zr solutions, where Zr acts as analogue for Pu (IV), mixed Mo-Zr species, being the precursors of the well characterized precipitate ZMH phases, form in solution. The relative abundance of the mixed Mo-Zr species with respect to Mo decreased with decreasing concentration of Zr in the sample. Mixed Mo-Eu species are formed in the Mo-Eu samples. For liquid-liquid extraction steps higher acidic strength is to be preferred since the species distribution is much simpler than at lower acidity. Experiments were also performed where thorium was the analogue for Pu(IV) and a similar behaviour was seen as shown in Figure 5.

The mass spectra of the different solutions show that Th forms monomeric and dimeric species [Thx(OH)a(NO3)b]+ (x = 1,2, a = 0–3 and b = 0–3). Mo forms pentavalent monomers and hexavalent dimeric, trimeric and lager polymeric species in accordance with earlier measurements of Mo alone, indicating that the formation of Mo species is not influenced by addition of Th in the solutions. Besides pure Mo species different mixed Mo-Th species [MoxThyOz(OH)a(NO3)b]+ (x = 1–3, y = 1–2 and z = 2, 5 and 8) form in the solution. A hypothesis test shows that part of the pure Th dimeric species might be artefact species. Mixed Mo-Th species [MoThO2(OH)2(NO3)3]+ and [MoThO2(OH)3(NO3)2]+ are most likely real solution species, because no Mo (VI) monomeric species exists in the solution which could form artefacts together with Th monomers in the same ESI droplet. These mixed species have up to 9.0 % relative abundance with respect to Mo and 3.4 % with respect to Th, respectively. The relative abundance of the mixed species decreases with decreasing Th concentration and increasing nitric acid strength in the sample with respect to both Mo and Th. For mixed Mo-Th species such as [Mo2ThO5(OH)3(NO3)2]+ at least partly formation during the measurement process cannot be fully excluded.
Selective separation of Mo by precipitation or selective extraction would facilitate the subsequent reprocessing process as well as the recovery of isotopically tailored Mo. After the Mo recovery initial process development studies for partitioning has been performed. In addition, the separation of remaining traces of radionuclides and stable impurities from the recovered molybdenum matrix solution prior its conversion to the metal and recycling has been studied.
Extraction using CYANEX® 600 shows some promise for the purification of the molybdenum matrix material of CerMet fuels. Further studies will be carried out with solutions containing actinide ions and with solutions containing Mo(VI) in concentrations as they would result from dissolution experiments.
Assuming the standard hydrometallurgical way of molybdenum reprocessing, the most significant potential (radionuclidic) impurities in the (ammonium) molybdate at the final stages of the process are expected to be caesium, strontium and technetium. KNiFC- PAN and Ba(Ca)SO4-PAN were identified as the most prospective materials for the chromatographic separation of caesium and strontium, respectively, from molybdenum solutions with relatively high molybdenum concentration (100 g·L-1 Mo) and slightly alkaline pH (pH 9.1). Similar experiments are ongoing for technetium. A final concluding experiment will be designed and performed aiming at the demonstration of the possibility to separate all the three radionuclidic contaminants in one step.
The thermal treatment of the fuel comprising the oxidation of molybdenum in air above 400°C and sublimation of the resulting MoO3 at 800°C is a promising method for the separation of the bulk molybdenum to simplify the dissolution and minimize the effect on the liquid-liquid extraction.
Neither a PUREX type process nor a DIAMEX type process extracts Zn(II) to a significant extent. As this is the case, Zn(II) will not interfere with SANEX type downstream processes.
The minor actinides (MA) and plutonium are the main contributors to long term radiotoxicity, heat load and proliferation risk of radiotoxic waste generated during nuclear reactor operation.

A lot of research nowadays is performed on the partitioning and transmutation of these nuclides. One of the methods under consideration to reduce the long term radiotoxicity of the remaining waste is burning the actinides in uranium free fuels in Accelerator Driven Systems (ADS).
Pure MgO reference samples as well as samples containing CeO2, as surrogate for TRU oxide, were fabricated and thoroughly characterized. Batch dissolution experiments were used for a systematic macroscopic study of dissolution kinetics. Additionally, the evolution of the MgO pellet surface during the dissolution process was investigated with a more sophisticated microscopic approach which allowed an estimation of the total surface area of the pellet and establishment of a surface normalized dissolution rate. MgO dissolves via a two-stage mechanism including the intermediate formation of a brucite-like surface layer, see Figure 6 and Figure 7.
The microstructural investigations showed a heterogeneous evolution of the pellet surface, which is attributed to the dissolution of different surfaces of a MgO crystal via heterogeneous etch patterns and to an inhomogeneous density of the pellets. Accordingly, some regions of the pellet dissolve fast but a pellet structure remains until complete dissolution. Apart from the surface evolution additional evidence for a surface controlled dissolution were obtained, such as an activation energy of 52 ± 8 kJ/mol. The dissolution of MgO is possible even under mild conditions (2.5 mol/L HNO3, RT). The average dissolution rates of MgO in HNO3 of different concentrations are rather similar, and the dissolution rate is strongly dependent on the dissolution temperature. The dissolution rate was normalized to the total surface area, which was obtained by considering the additional surface area due to the formation of holes. The additional surface obtained by considering the development of holes in the pellet resulted in a dissolution rate of approximately 1.2·107 mg·m-2·d-1. The dissolution of the MgO/CeO2 pellets shows that during dissolution of MgO, the bulk CeO2 remains undissolved allowing a rough separation of MgO and CeO2. In none of the experiments more than 1% of CeO2 was dissolved. A dissolution behaviour similar to that of CeO2 is expected for PuO2 so that MgO could be separated in a first dissolution step at mild conditions, while the PuO2 would be dissolved in an additional step utilizing e.g. boiling concentrated HNO3 and Ag(II) as catalysts. However, an overall evaluation of the dissolution behaviour of IMF can be given only if the effect of irradiation is considered. Therefore, dissolution experiments with irradiated targets are necessary.
1.3.2.3 WP2.2 Basic studies on oxide fuels for Generation IV systems
Oxide fuels for Generation IV systems and IMF can contain high contents (up to 50%) of plutonium and minor actinides. However, the adverse effect of the MA-content (specifically Am) on the dissolution capability is unknown. Yet this effect forms a crucial aspect in the design of a reprocessing flow-sheet for oxide fuels for Generation IV systems, as well as for Inert Matrix Fuels. It was the objective of this work package to study, generally, the basic dissolution properties of oxide fuels with a high concentration of minor actinides. This enables establishing a relationship between Am-content and dissolution capability. This addresses open issues concerning the maximum achievable Minor Actinide content in fuels and targets for transmutation purposes.
The composition of irradiated (U,MA)O2 blankets and (Pu,MA)O2-Mo inert matrix fuels has been calculated with the Serpent Monte Carlo code. The resulting transmutation rate of minor actinides in the MABB case is 15% lower than previous calculations made by CEA using the ERANOS code system. The reason for this discrepancy will be investigated in an update of the present report. The predicted plutonium oxide fraction in the MABB pellets by the end of irradiation is less than 8.0 percent, which should not pose a serious problem for the dissolution of this material, even taking the above discussed discrepancy into account.
In the case of the inert matrix CERMET fuel, the calculated transmutation rates of minor actinides are slightly higher than for the MABB. The reason for this is the higher neutron flux present in the IMF assemblies. The calculated fraction of plutonium in the oxide phase of the fuel at EOL exceeds 40%, which may constitute an issue for the dissolution performance. Dissolution tests of (Pu,MA)O2 samples with a Pu fraction of up to 50% therefore should be carried out to assess the solubility limit of this fuel composition.
A generic problem related to the heterogeneous transmutation approach is the high alpha-decay heat pertaining to the inventory of 242Cm, 244Cm and 238Pu. The IMF fuel appears to be somewhat more vulnerable to this issue than the MABB concept. Possible solutions to address this problem include carrying out transmutation of neptunium in homogeneous mode, decreasing the number of fuel pins per assembly, or reducing the residence time of the fuel/blanket assembly dedicated to transmutation.
Based on the calculation two types of pellets have been fabricated by powder metallurgy method and characterized. The composition of the pellets is 28.4 wt% AmO2 – 71.6 wt% PuO2 and 52.4 wt% AmO2 – 47.6 wt% PuO2. The pellets have density of about 75 and 71, resp. %TD, see Figure 8.

The composition of these unique pellets was investigated using XRD as shown in Figure 9.

1.3.2.4 WP2.3 Conversion from solution to oxide precursors
Different sol-gel methods have been studied for fabrication of microspheres. The influence of the gelification agents HMTA and urea on the gelation temperature was studied for U/Nd containing mixtures (5, 25 and 40 % Nd content) at Forschungszentrum Jülich. It could be shown that an increase of the neodymium content in the U/Nd mixture leads to a gelation at a lower temperature. The maximum temperature for the composition containing 5 % Nd was ≈ 55°C, while the temperature recorded for the 40 % Nd mixture was ≈ 40°C (R(HMTA) = 1.0, R(urea) = 2.0). The amount of HMTA present in the sol has a huge contribution to the gelation temperature, independent on the urea concentration, a smaller gelation temperature was recoded with an increasing R(HMTA) value. XRD measurements performed on the gelation products of the sample with R(HMTA) = 1.0, R(urea) = 1.6 and R(HMTA) = 2.0, R(urea) = 1.6 indicated the same phases. An increase of the urea concentration has no significant influence on the gelation temperature for R(HMTA) ≤ 1.4. For the experiments with R(HMTA) = 1.6 and 1.8 the gelation temperature increases with an increasing R(urea) value. Some composition had a gelation temperature below 13°C (initial condition) and no temperature could be recorded. This study showed that an adjustment of the gelification agents, depending on the U/Nd ratio, has to be done to ensure a constant gelation temperature during the synthesis of different U/Nd compositions via internal gelation.
Moreover, it was demonstrated, that the preparation of UO2/Nd2O3 with Nd contents up to ≈ 45 % is possible by the internal gelation technique. ADUN solutions were used as precursor for the fabrication. During the synthesis constant conditions regarding the gelification agents urea and HMTA were kept (R(HMTA) = 1.35, R(urea) = 1.8). U/Nd green bodies in a χ(Nd) range of 0 − 42.63 % were successfully synthesised. A spherical particle shape was proven by SEM and with optical microscopy. The thermal behaviour of the U/Nd microspheres was investigated by TG/DSC, in the region of 550-700°C the particle compositions with χ(Nd) ≤ 17.40 % transit endothermic accompanied with an stepwise mass loss. The spheres containing higher Nd amounts showed an exothermic effect. Moreover, an endothermic effect at > 900°C was observed for some samples. XRD analyses show orthorhombic and cubic crystal lattice structure of the particle compositions treated in air (θ = 1300°C), which agrees with SEM investigations. The lattice parameter of the cubic structure was determined. These calculations were done for the particles treated in H2:Ar atmosphere at 1300°C as well. The expected linear trend according to Vegard’s rule was confirmed for compositions containing up to 27.59 % Nd. Particles containing more Nd behave unexpected and consist of 2 phases. The particles sintered at 1600°C behave more or less the same up to 25 % Nd, but in this case, compositions containing more Nd show only one cubic phase. However, they have a deviation on the expected linear behaviour. The results of this work show that equilibrium solid solutions of the sensitive UO2/Nd2O3 system can be fabricated by the internal gelation synthesis route with Nd contents of ≤ 42.63 %.
The synthesised particles were analysed by X-ray powder diffraction, the presence of a fluorite structure for the kernels sintered at 1300°C and 1600°C in Ar:H2 was found. The introduction of a trivalent ion into the UO2 lattice causes a charge non-equilibrium. This can be compensated by the formation of oxygen vacancies, the oxidation of U(IV) to U(V) or even to U(VI). The data of our particles indicate a charge compensation by formation of oxygen vacancies, the radius of the vacancy was determined to be rov = 1.11 ± 0.02 Å. The value is within the confidence level of published data, but more precise. The lattice distortion was used to calculate a “corrected” lattice parameter. This correction leads to a linear depending lattice parameter on the neodymium content for all investigated compositions treated at 1600°C. For the particles annealed at 1300°C the linear depending single phase region extended up to a Nd content of 33 %. The lattice parameter in the two phase region is constant and independent on the composition. Moreover, theoretical lattice parameters were calculated for each charge compensation model, and a good correlation of the corrected experimental data with the theoretical values was observed.
Another method of mixed oxides U1-xNdxO2 preparation has also been proposed based on sol-gel processes. First step was preparation of ascorbate uranyl-neodymium sol in the system UO3 - ascorbic acid - neodymium nitrate – ammonia - water. Sols were gelled to microspheres (diameter <200 μm) in processes elaborated in INCT; this process involves the extraction of water from emulsion drops formed in 2-ethylhexanol-1. It was observed that further thermal treatment of those gels required extremely slow heating rate because presence of ammonium nitrate. In order to avoid this inconvenience Double Extraction Process was elaborated, based on the simultaneous extraction of water and nitrate ions by 2-ethylhexanol-1 contained (1-2 vol.%) Primene JM-T.
After carefully selected parameters of calcination and reduction, selected also using thermogravimetric studies, perfectly spherical particles ∅<150 μm) of U1-xNdxO2 (x = 0.1-0.4) were obtained. Larger spherical particles (kernels) have been prepared by internal gelation method from described sols with addition hexamethylenetetramine (HMTA). This reagent has generated of ammonia into drops by heating. As indicated results of SEM, EDX, XRD studies in all final products only solid homogenous solutions (U1-xNdx)O2 (x = 0.1 – 0.4) in the form of a well-crystallized single-phase fluorite structure was observed. However for samples prepared using gelation by water and nitrate extraction, concentration of neodymium is slightly higher in the centre of particles than in the surface. This phenomenon was not studied in kernels prepared by internal gelation. The lattice parameter (a) decreases linearly with neodymium content till 30 mol.% of neodymia. Explanation of this phenomenon has been presented on the basis of three factors, namely ionic size of the dopant, doping of cation with a lower valency (Nd3+) and dependency the Nd content with vacancies in the anionic sub-lattice.
Presented results clearly show that describe sol gel processes - using ascorbate sols - can be successfully apply for fabrication of U1-xNdxO2 in form of irregular particles as well as spherical, in diameter range 1-2000 μm.
Maximization of UO22+ and Nd3+ adsorption by two different weak acid cation exchange resins, Amberlite IRC-86 and Lewatit TP-207, followed by thermal treatment in different atmospheres at different temperatures to produce microspheres has been studied. The work is based on the weak acid resin (WAR) process and neodymium is used as surrogate for trivalent actinides.
The WAR–process can be used for the preparation of UO2 based particles containing Nd2O3 in the form of a solid solution. A spherical shape was kept through the process in the case of the ion exchange resin Lewatit TP-207. Amberlite IRC-86 resulted with a huge amount of broken particles and is not suitable for production of microspheres.
Due to the difference of adsorption kinetics of Nd3+ and UO22+ the optimum contacting time was found to be 18 h to maximize loading while the molar ratio of Nd3+ adsorbed is equal to the molar ratio of Nd3+ in the initial solution. The temperature but the adsorption increases with increasing pH and maximizes at a pH of about 3.
Treatment in oxidizing atmosphere (air) at 1173 K followed by a further treatment in reducing atmosphere (5 % H2/95 % Ar) produced microspheres with a homogeneous mixture of UO2 and Nd2O3. The particles have been investigated using SEM, EDX and XRD analysis.
Further measurements are required on the influence of the initial ion concentration on the adsorption. Kinetic measurements using comparable initial circumstances (such as pH) and at various temperatures can be performed to increase the adsorption of Nd3+ and UO22+. Furthermore, comparable studies with Am(III) and Cm(III) could be performed to determine if the results obtained for Nd(III) are representative for trivalent actinides.
The verification of the „proof of principle” of the possibility to prepare oxides of selected actinides incorporated into ceramic matrices like MgO using PAN–process has been performed. In model experiments, Ce and Eu were used as surrogates for Pu and Am.
The PAN–process proposed can be used for the preparation of CERCER-type pellets with MgO matrix that contain mixed oxides of Ce and Eu in the form of their solid solution. The density of sintered pellets is in the range of 80–90 % of the theoretical density (TD). It would be worthwhile to verify the possibility of pellets preparation from small sintered beads (dimension after sintering ˂ 0.025 mm or ˃ 0.1 mm) by mixing with MgO, pressing, and another sintering. Composite microdispersed or macrodispersed fuel with MgO matrix should be obtained.
Relatively strong CERCER-type beads with MgO matrix incorporating Eu or Ce oxides or their mixture can be produced by infiltrating MgO–PAN beads with Eu, Ce, or Eu+Ce solutions. The prepared beads contained 25–30 % (w/w) of the oxide(s); the contents of up to 50 % (w/w) should be achievable by single or maximum double infiltration. If smaller sintered beads are prepared, better homogeneity and smaller porosity is expected to be achievable.
The photo- or radiation-induced preparation of nanocrystalline UO2, ThO2 and solid (Th,U)O2 solution and their subsequent pelletizing has been studied. The preparative method was based on formation of amorphous precursors in aqueous solutions containing uranyl and/or thorium nitrate (0.005–0.01 mol·dm-3) and ammonium formate (0.1 mol·dm-3) by mercury lamp (low or medium pressure), accelerated electrons or gamma irradiation. The obtained precursors were carbon free, allowing the direct heat treatment in reducing atmosphere without pre-treatment in air. The EXAFS analyses of the amorphous precursors indicated that the Th-containing precursors were ThO2 (direct formation of ThO2), whereas the U-containing precursors were a mixture of U(IV) and U(VI) compounds.
The heat treatment of amorphous precursors under various atmospheres at minimum temperatures in the interval of 300–700°C yielded nanocrystalline UO2, ThO2 or solid (Th,U)O2 solutions with high purity and linear crystallite size from 3 to 51 nm.
The prepared nanocrystalline oxides were compacted into pellets using stearic acid as a die-wall lubricant, without any binder, at a pressure of 500 MPa. The green pellets were subsequently sintered at 1300°C under a 20Ar:1H2 mixture (UO2 and (Th,U)O2 pellets) or at 1600°C in ambient air (ThO2 pellets). The densities of green pellets reached values between 36 and 57 % of theoretical density, whereas sintered pellets reached values from 90 % to 97 % of theoretical density. The prepared pellets were highly durable as only one pellet was typically sufficient for characterization and sintering procedures without substantial damage to the pellet.
Photo- or radiation-induced preparation of nanocrystalline UO2, ThO2 and (Th,U)O2 oxides presents a convenient preparation technique based on simple preparation of solution, its irradiation, separation of formed precursor and its heat treatment. This preparation technique is conveniently dust-free and presents an alternative to the powder mixing technology, due to the fact that a solid (Th,U)O2 solution can be prepared. In comparison to the sol-gel process, detailed control of the process is not necessary, which represents an advantage for industrial synthesis of these materials. Based on obtained results, the photo- or radiation-induced preparation methods offer promising alternative routes for the preparation of high density pellets of (mixed) oxide nuclear fuel.
Powder Eu2O3, CeO2 and their mixed oxides with uranium Eu2O3-UO2 and CeO2-UO2 were prepared by the UV or e-beam irradiation of aqueous solutions containing metal nitrates and ammonium formate. Irradiation resulted in the formation of solid precursors, which were subsequently separated from the solution and heat treated at moderate temperatures to obtain pure crystalline oxides. In the case of mixed oxides, formation of solid solutions was observed. It was also found that the presence of UO2 strongly affects the morphology of prepared oxides.
The composition of solid Ce-precursors formed under irradiation is dependent on the composition of initial solution and the source of radiation.
1.3.3 Domain 3 Nitride fuels
1.3.3.1 Introduction
Nitride fuels are considered for industrial application in both fast reactors, light water reactors and heavy water reactors, due to the higher density of the fuel, leading to better conversion ratios and longer residence time. In the context of minor actinide transmutation, the high thermal conductivity of nitride fuels, in combination with a high failure temperature, results in larger margin to fuel failure. Therefore one may load fuels with higher fractions of minor actinides, increasing thus the achievable minor actinide burning rates.
Historically, UN was used as driver fuel for two core loads in BR-10. Otherwise, the experience is limited to test irradiations of a few hundred fuel pins for space reactor applications and fast reactor programs in the US, EU, Japan and Russia.
These test irradiations have had a mixed outcome, with failures relating to chemical instability, high swelling rates and severe PCMI leading to a waning interest in this fuel composition by the mid 90’s.
The excellent performance of (U,Pu)N and (Pu,Zr)N fuels in test irradiations conducted at moderate linear ratings have led to renewed interest in this fuel type. In particular, a decision was taken to build and operate the lead fast reactor BREST-300 with (U,Pu)N fuel. To this end, several experimental (U,Pu)N assemblies have been fabricated by Siberian Chemical Combine (SCC) for irradiation testing in BN-600. A large scale manufacturing plant is currently under construction in the Tomsk region, with the aim to produce mixed nitride fuel on a commercial basis in 2018.
In Sweden, KTH, Chalmers and Uppsala University have proposed to build ELECTRA-FCC, a low power lead-cooled reactor with associated fuel cycle facilities. The foreseen fuel for ELECTRA is (Pu,Zr)N. The main function of ELECTRA in the European context would be to function as an education and training facility.
Questions arising when studying the feasibility of applying nitride fuels include the dissolution rate of inert matrix fuels in nitric acid, the manufacturability of pellets with sufficiently high density, and the cost of 15N required in order to avoid excessive neutron absorption leading to the production of 14C.
Within the ASGARD-project, the following activities on nitride fuel were carried out within Domain 3 to investigate the feasibility of applying nitride fuels in a closed fuel cycle:
1) Dissolution tests of fresh and irradiated fuel
2) Manufacture of UN, PuN, (U,Zr)N and (Pu,Zr)N fuel using wet and dry routes
3) Improvement of 15N economy in terms of enrichment and savings during manufacture and recycle.

1.3.3.2 WP3.1 Dissolution of fresh and irradiated nitride fuels
1.3.3.2.1 Fresh fuel
Prior to ASGARD, it was known that UN powders and pellets dissolve rapidly in nitric acid even at room temperature. Within the project, the impact of impurity and porosity on the dissolution rate of UN-pellets was investigated systematically for the first time. The major effect was shown to derive from the amount of open porosity of the pellet. For samples with ≈ 1000 ppm of oxygen and carbon impurities, the time to reach full dissolution differed by an order of magnitude between a pellet with 7.5% open porosity (1 hour to reach full dissolution) and a pellet with 1.7% open porosity (10.5 hours to reach full dissolution). It could be noted that since the dissolution mechanism is exothermic, the reaction was not occurring at constant temperature, in particular for the sample with high porosity.
Dissolution rates of (U,Zr)N pellets in nitric acid were found to be negligible at room temperature. Even at 110°C, dissolution did not reach completion. The samples tested contained some initial phases of UN and ZrN not in solid solution. Post-test analysis revealed a preferential dissolution of the pure UN phases, indicating that formation of a complete solid solution of (U,Zr)N will substantially reduce dissolution kinetics.
Two types of unirradiated (Pu,Zr)N materials were used for dissolution experiments in ASGARD, where one was in a powder form (deriving from the CONFIRM project) and the other in the form of (Pu,Zr)N/C microspheres, allowing to evaluate the effect on surface area as well as carbide content.
For the CONFIRM powder, full dissolution required either a combination of high temperature and high acidity (80°C for 10M HNO3 or 110°C for 4M HNO3), or the addition of catalytic amounts of hydrofluoric acid at room temperature in 10M HNO3. No difference in dissolution rate of Pu and Zr was observed.
The (Pu,Zr)N/C microspheres used for dissolution tests were produced via the internal gelation process. Before dissolution, the microspheres were sintered under argon during, 5 hours at 1700°C in order to mimic the dissolution behaviour of a sintered pellet. SEM-mapping of a cross-section of a sphere revealed an inhomogeneous microstructure of the material, consisting of one zirconium rich phase and one plutonium rich phase.
As for all inert matrix materials, the sintered (Pu,Zr)N/C microspheres were unaffected by exposure to cold nitric acid even at very high (14M) acidity. When the temperature is raised to 120°C dissolution slowly takes place. After > 27 h a black solid residue still remained, contrary to the dissolution of the fresh CONFIRM powder. Upon addition of HF, complete dissolution was achieved and by totally dissolving three different microspheres the zirconium and plutonium content was found to be 58.5 ± 1.6% Zr and 41.3 ± 1.4% Pu. As the solution from the dissolution in hot nitric acid contained circa 80% Zr and 20% Pu, the black solid residue should be rich in plutonium. When adding HF, plutonium is observed to dissolve more quickly than zirconium, in difference to the data for CONFIRM powders.
1.3.3.2.2 Irradiated fuel
In the FP5 CONFIRM project, two (Pu0.3,Zr0.7)N rodlets were irradiated in the HFR reactor, reaching an average burn-of 10% fission in actinides. Within ASGARD, five pin slices were prepared for dissolution tests. The dissolution was carried out in boiling nitric acid (110°C) with and without addition of hydrofluoric acid. In one case, a silver catalyst was added.
In the test using pure nitric acid, the central part of the pellet was dissolved, leaving an undissolved residual layer at the high burn-up rim of the pellet. Adding HF, most of the pellet is dissolved, whereas black particles are observed in the solution which is similar to what was observed for the sol-gel non irradiated fuel described above. WDS/EDS analysis indicated that these particles contain zirconium, possibly a mixture of ZrN and ZrO2. The silver catalyst did not appear to improve the dissolution performance which is unexpected in the case of formation of a PuO2 phase which usually readily dissolves under these conditions.
1.3.3.3 WP3.2 Fabrication of nitride fuels
Two major classes of nitride fuels are considered for application in closed fuel cycles: (U,Pu,MA)N as driver fuel in Generation IV reactors, and (Pu,MA,Zr)N inert matrix fuels for minor actinide burning in Accelerator Driven Systems. In the former, 80-85% of the fuel consists of UN, whereas in the latter, > 60 volume percent of the fuel is expected to be ZrN in solid solution with the actinide nitrides. Uranium nitride for Gen-IV systems may be manufactured using power metallurgy, but plutonium and americium nitrides are preferably synthesised using wet routes.
For manufacture of UN powders, hydriding/nitriding of metallic uranium was applied. In conjunction with spark plasma sintering this approach allowed to produce pure (< 600 ppm oxygen, < 1000 ppm carbon) and highly dense UN pellets (up to 99.9% TD). The influence of carbon and oxygen impurities on dissolution could be investigated by mixing pure UN powders with the desired fraction of UO2 and or UC prior to sintering. In such end products, UO2 appeared as a separate phase, whereas UC formed a solid solution with UN. Porosity could be controlled in the range from 0 to 12% by adjusting temperature and applied pressure during sintering, as illustrated in Figure 10.

For the purpose of producing of (U,Zr)N, zirconium nitride powder was synthesised from zirconium metal by the hydriding/nitriding method. Here, the intermediate zirconium dihydride product must be milled prior to nitriding. Milling in air introduced resulted in a significant fraction of zirconium oxide being present in the end product.
Co-milling impure ZrN with UN using a 316 stainless steel milling cup and milling balls was conducted. Mixed nitride pellets produced by SPS from such powders turned out to have a dual phase micro-structure. In order to achieve solid solution (U,Zr)N, using a tungsten carbide (WC) milling cup and milling balls for the co-milling process was found to be an adequate approach. (U,Zr)N pellets for dissolution tests were made from dual phase powders (solid solution powders became available only by the very end of ASGARD) and contained between 8 to 22% light atom oxygen, and between 1 and 22% light atom carbon. Whereas the total porosity fraction could not be determined, the fraction of open porosity varied between 0.9 and 10.5%. PuN pellets were produced through carbo-thermic nitriding of PuO2 powders, followed by pressing and sintering under nitrogen atmosphere in a graphite furnace installed in a glove box. Sintering temperatures of up to 1800°C were applied. No oxide phase was detected in the end product and the PuC content in solid solution with PuN was determined from XRD lattice parameters to range between 600±500 to 3100±900 ppm.
The pellet density derived from geometrical measurement of pellet dimensions varied from 63 to 79% TD.
Gamma spectroscopic measurements were performed on the powder batches produced, on the starting oxide powder and one of the dissolved sintered pellets to examine the degree of americium losses from the material during carbo-thermic nitriding and sintering. An interesting conclusion is that a carbo-thermal nitriding time of 4 hours at 1400°C and 2 hours in between 1600 and 1700°C is not enough to result in Am losses large enough to be accurately determined by the procedure used. However, losses during sintering of pellets are likely larger. The total Am loss measured after dissolution of the pellet sintered at 1800°C was estimated at 5-17%.
In order to mimic the real situation where the nitride fuel is actually made from dissolved and separated used nuclear fuel pure PuN pellets were also made using the sol-gel technique. This is either the first or second time in the world this material is produced by this technique. The final investigation of the produced pellets not yet performed but there is no indicating that the material is very different from the material made when starting with the pure PuO2. Naturally this is highly interesting for recycling purposes.
(Pu,Zr)N microspheres were prepared by the sol-gel route. Several different methods for plutonium sol preparation were investigated. First, zirconium was added to the solution to obtain a metal feed composition of 40 mole% Pu and 60 mole% Zr. Zirconium was added as ZrOCl2. The Pu and Zr containing solution was then pH adjusted by addition of 25 w% aqueous ammonia to reach sufficiently low H+ concentration for gelation by HMTA addition to be possible. Use of NH4OH was considered the preferential method for pH elevation. The concentration of Zr and Pu in the solution was adjusted by dilution using both aqueous ammonia and MQ water if needed. Urea and HMTA to metal ratios were varied in order to find suitable gelation parameters.
Triton X-100 and carbon was added separately to the sol after pH elevation had been completed. Washing of the produced microspheres were performed in petroleum ether and NH4OH (aq), followed by ageing in NH4OH (aq).
Carbothermal nitriding of the carbon containing microspheres was performed in N2 + 5% H2 gas by heating the sample to 1400°C. Subsequently, de-carburisation was carried out at 1650°C. The microspheres thus produced were estimated to be about 600-750 μm in diameter and the surface of the microspheres showed cracking, as seen in Figure 11. The cracking could be caused during the carbothermal nitriding either when residual water and gelation chemicals vaporise or when CO is formed during nitride formation and escapes from the microspheres. If the aim is to produce pellets from the microspheres, some degree of porosity is advantageous since it is beneficial when crushing microspheres during pelletisation if one wants to avoid so called “blackberry” structure.
The resulting microspheres were pressed into pellets and sintered under nitrogen or argon at 1700°C. SEM characterisation of sintered pellets revealed no remains of blackberry structure in the material. EDX measurements on large areas of the pellets surfaces indicated plutonium content of about 30 – 40 % in the pellet sintered in argon while the pellet sintered under nitrogen displayed regions of almost only plutonium and very little zirconium. XRD analysis showed that the pellets were carbonitrides rather than pure nitrides, although the carbide fraction was difficult to determine from XRD analysis alone.
The carbo-thermic nitriding method for producing pure UN from UO2 requires that CO gas is removed from the reaction furnace, making it difficult to retain expensive 15N gas. Within ASGARD, the conventional hydriding/nitriding process form producing uranium nitride from metallic uranium was modified in order to permit a final reaction step where the consumption of nitrogen gas in the reaction chamber could be made near complete without the need for removing gaseous reaction productions. The modified process proceeds as follows:
1) React metallic uranium with hydrogen, to form a fine powder of UH3.
2) Raise the temperature to dissociate UH3 into a fine powder of metallic uranium and hydrogen gas
3) Nitriding the fine powder of metallic uranium to form sesquinitride
4) Raise the temperature to dissociating the sesquinitride into UN (while recovering nitrogen).
Steps 1-3 were conducted as part ASGARD, showing that near complete uptake of nitrogen gas can be achieved while keeping reaction temperatures at or below 600°C.
1.3.3.4 WP3.3 Enrichment of nitrogen in 15N
1.3.3.4.1 Recovery of 15N
Hydrolysis of fresh UN pellets was carried out in steam at a partial pressure of 0.5 bar in order to investigate the feasibility of recovering N-15 in the form of ammonia following the reaction (see Equation 1).
The rate of conversion was found to depend on porosity of the pellets as well as on temperature of the steam. For a dense pellet (96% TD), near complete conversion was achieved after seven hours of exposure at 500°C, whereas at 425°C, the majority of a pellet with a density of 98% TD remained intact after an exposure of five hours. A highly porous pellet (78% TD) was fully converted under vigorous release of hydrogen at an exposure temperature of 425°C already within two hours.
1.3.3.4.2 Required 15N fraction in fuels
Both oxygen and nitrogen contribute to the production of 14C in nuclear reactors. Table 2 shows the calculated spectrum averaged cross sections for 14C production in the fuel of an LFR operating with (U,Pu)N fuel. The largest cross section is found for 17O(n,α)14C, which consequently is the most important contributor to 14C production in fast reactor oxide fuels. The production cross section from 15N is 300 times smaller than from 14N, which means that above 99% enrichment, the 15N(n,d)14C generation path starts to become of significance in fast reactor nitride fuels. The data are compared with production rates in a PWR spectrum.
Assuming a nitrogen concentration of 15±10 ppm in an LWR fuel, leads to a total 14C production rate of 7.5±1.4 g/GWe x year in LWRs, out of which 5.6±1.4 g is present in the spent fuel.
The production of 14C in a typical LFR was calculated adopting a 15N enrichment of 99%. The resulting inventory at EOL is 62 grams, corresponding to a production rate of 29 gram/GWe x year. Since fast reactors operating on nitride fuel in a closed fuel cycle would constitute less than 20% of the fleet, one finds that 99% enrichment is sufficient for the fast reactors to produce less 14C than the light water reactors.
1.3.3.4.3 Reducing cost for 15N production
The market price for enriched 15N today exceeds 100 €/gram and is prohibitive if nitride fuel is to be produced using this isotope, as required if 14C production in fast Generation IV reactors is to be maintained below the rate of generation in LWRs. The industrial process foreseen to be used is the NITROX system, based on chemical exchange between nitrous oxide and nitric acid according to Equation 2
characterised by a single stage separation factor of α = 1.055 for a 10M HNO3 solution at 25°C. Within the ASGARD project, the production of 15N in a pressurised column was investigated, in order to permit an increase in flow rate of nitric acid, and hence in production rate of 15N.
In the NITORX process, nitrous oxides are produced from nitric acid using sulphur dioxide according to Equation 3.
By operating a NITROX separation column at an overpressure of 0.8 atm, it was shown that the flow rate of nitric acid could be increased by 60%. The 0.8 m length product refluxer of the experimental separation plant was constructed of hastelloy, a material which is resistant to corrosion of both nitric and sulphuric acid. The operation of 15N cascade at 0.8 atm, instead of atmospheric one, will allow 1.57 times increasing of the nitric acid flow rate and of the 15N production of a given plant. The volume of a cascade for 15N production, operated at 0.8 atm pressure instead atmospheric one, will be 1.59 times smaller than the volume of a plant operated at atmospheric pressure. The capital investment for a new separation plant will decrease in the same proportion. Moreover, since sulphur dioxide, an expensive chemical, is used in the NITROX process, a process for converting sulphuric acid (the waste product of the NITROX process) into SO2 using various catalysts was investigated. It was found that ferrous oxide (Fe2O3) could perform equally well at high temperature (850°C) as more expensive platinum or palladium/alumina catalysts.

1.3.4 Domain 4 Carbide fuels
1.3.4.1 Introduction
Carbide fuels have been considered as fuels for Fast Reactors since the 1960s and many irradiation experiments on U and U/Pu carbide have been performed globally, with recent successes particularly for the Fast Breeder Test Reactor at Kalpakkam, India. The physical properties of carbide fuel make them attractive because they are conducive to high specific rod powers with relatively low fuel centre temperatures: start of life power capability is increased, power-to-melt margin is increased and fatter (more economic) pins are facilitated. As fuel temperatures are low, the fuel suffers little or no restructuring and the release of fission gases and volatile fission products is low. Under transient conditions, high thermal conductivity and low thermal expansion offer certain advantages, i.e. lower fuel/clad mechanical interaction (FCMI) levels in over power transients and lower cladding temperatures in under cooling events. Operationally, better internal breeding reduces reactivity decline over the fuel lifetime, which eases control rod management.
Carbide fuel chemical properties provide the advantage that under normal operating conditions fuel/stainless steel clad chemical interaction is virtually absent; additionally, swelling due to fuel/coolant interaction in a failed pin should not occur to the benefit of minimising contamination of the reactor circuit. Thus two sources of concern present with oxide pins are potentially eliminated.
On the down side there are a number of evident problems with carbide fuel. Performance limitations are of concern because of the significant potential for unacceptable fuel/clad mechanical interaction (FCMI) due to the high swelling propensity from gases trapped within the fuel matrix. This would lead to the possibility of extensive reaction between fuel and cladding at higher transient temperatures, thus limiting the carbide fuel centreline pellet temperature and the overall economic performance of the reactor. It is then clear that if the fuel swelling mechanism can be avoided/reduced, carbide fuel could have significant advantages towards FCMI compared to oxide fuel.
In the fuel cycle, manufacturing involves the handling of pyrophoric powders, which may not be amenable to scale-up; additionally, fuel reprocessing is problematic since mixed U/Pu carbide dissolution results in undesirable side products in nitric acid.
Domain 4 of the ASGARD project aimed to seek solutions to the problems of fuel swelling and address the issues concerning the reprocessing of carbide fuels. The objectives were to fabricate carbide fuel in such a way that avoids fuel swelling and identify a suitable route for reprocessing of carbide fuels. In addition, more information was gathered on the pyrophoric nature of carbides and novel methods for carbide formation were explored.
1.3.4.2 WP4.1 Design and manufacture of carbide fuels for Generation IV FR systems
In this section, a number of issues were addressed with the assumption that (U,Pu)C fuels would be either formed by conventional cold press and sintering methods or by using a range of sintered microspheres made by microwave gelation of a uranium and plutonium precursor. Initially modelling was performed to establish a suitable target pellet geometry and optimal sphere diameter mix for the two approaches. This was then followed by experimental work to produce the materials.
1.3.4.2.1 Pellet pressing and sintering
The mandate of this task is to establish the design of carbide fuel (both pelleted and sphere-pac designs) that is optimised in terms of reducing the fuel swelling whilst maintaining good thermal properties. This work focuses on the development of (U,Pu)C pelleted fuel pins and reports the generation of a reference (U,Pu)C pin design, the development of a modelling methodology and the optimisation of the design. A licensing study also features as part of the optimised pin design assessment. Specific diverted effort was used to assess the viability of enhancing the fuel creep with additives and radial density gradient and their influence (if any) on the fuel swelling properties. The effect of nickel as a creep additive was studied and was found to reduce the peak clad hoop stress during fuel/clad mechanical interaction (FCMI) at levels of 0.1 wt% nickel. Consequently, the optimised pin design includes the effects of 0.1 wt% nickel on the fuel secondary thermal creep strain rate. Key phenomena were only sensitive to variations in the porosity gradient during FCMI. The maximum clad hoop stress appeared to be slightly sensitive to the porosity gradient and pointed to having a higher porosity in the outer region of the fuel. However, the stress results were highly scattered and rendered any postulation on the effect of a radial porosity gradient difficult to justify.
Consequently, the optimised pin design does not feature a porosity gradient. During the course of this work, it was discovered that the fuel swelling is most sensitive to the fuel temperature, which is characterised by the sensitivity studies on the peak mass rating, fuel pellet and cladding dimensions. The optimised pin design has a reduced fuel swelling and higher fuel temperatures whilst retaining the same discharge burnup as the reference pin design. Ultimately, this work conforms to the ASGARD task mandate of reducing the fuel swelling whilst maintaining good thermal properties.
Further to the design optimisation processes, the optimised pin design was subjected to steady-state design licensing criteria on clad strain, clad corrosion and fuel melting. The optimised design was scrutinised and found to conform to all the licensing limits.
1.3.4.2.2 Production of uranium carbide pellets
Significant infrastructure investment was made with the installation of a carbide fuel fabrication line. The line included a press with hold down function and a graphite sintering furnace capable of reaching 2000°C. The pellets were successfully produced to give duel density and enhanced nickel content and could be produced to a density of >90% through sintering at 1750°C for 2 hours. Together with investment from NNL there is now an inert atmosphere pellet production line available for making exotic fuels for future projects. The results of this study provided an input into further reactor code modelling that was updated.
1.3.4.2.3 Production of sphere pac carbide fuel
The benefits of producing sphere-pac fuel lie in a reduction of carbide powder handling, which is pyrophoric, and the potential direct formation of dense microspheres of carbide that are simply added together in different size fractions to generate packing densities of > 80% of the carbide theoretical density. Compared to conventional solid press and sintered pellets of carbides, the centreline temperatures are slightly higher and the diameter of the pins needs to be slightly wider to account for the lower density of fissile material. However, the centreline temperatures are much lower than that of oxides of the same power rating and therefore some benefits of thermal behaviour remain, see Figure 12.
A relatively mature technology proven to work at laboratory scale in the past was applied to microwave assisted internal gelation. Applying this method to carbide production involves the formation of small droplets of a metal and carbon containing solution that pass through a microwave heated area to induce gelation (hydrolysis) of the metal solution forming a precipitation that is dehydrated during heating but which contains intimately mixed elemental carbon. The dried spheres can then undergo carbothermic reduction where the microspheres are heated to high temperatures (>1700 - 1900°C) converting the oxides to carbides. The microwave part of the process omits the requirement for time consuming decontamination of the product from silicon oil that is used in conventional gelation techniques.
The equipment to make uranium and plutonium carbide fuel microspheres by microwave induced internal gelation was not fully commissioned actively during the course of this project. The instrument was demonstrated to work inactively, producing very good microspheres of cerium oxides showing good size control see Figure 13.
1.3.4.2.4 Novel carbide production routes
As potential alternatives to conventional dry and gelation methods of carbide production,
the following technologies were explored:
· Colloidal formation of nano-crystalline carbide pre-cursors – where uranium chloride solutions are mixed with methylcellulose, which is the molecular carbon source. The addition of ammonia solution increases the pH and causes precipitation of uranium oxo-hydroxo nanoparticles stabilised by the methylcellulose. The resultant precipitate is heat treated (1000°C, Argon) to give an intimately mixed powder with nano-crystalline uranium oxide and element carbon. Carbothermic reduction then occurs at low temperature (1450°C, Argon) to give uranium carbide.
· Weak Acid Resin - method of loading Nd as a simulant for Am onto a resin and then using the carbon framework of the resin as an in situ source of carbon for carbothermic reduction. Experiments have shown that temperatures of greater than 1600°C will be required to get this reaction to go to completion.
· Polyacrylonitrile (PAN) beads – A range of beads loaded with varying complexants were used to retain actinides on the beads and then decompose the beads that act as an in situ source of carbon. The carbon containing metal oxide is then heated to 1700°C to achieve carbothermic reduction.
· Uranyl ascorbate – this process takes advantage of the solubility of UO3 in ascorbic acid to generate concentrated uranium compounds that could be combined with other actinides. The process was demonstrated to work with uranium to generate microspheres of carbides.
(Equation 4)
Each of these methods were pursued in order to reduce the temperature of carbothermic reduction and minimise the amount of milling required to ensure intimate mixing of the oxide and carbon powder or ensure the reactive surface area was high. Another benefit of a reduced temperature for carbothermic reduction is to minimise the potential for actinides, particularly minor actinides, to volatilise. All of these novel methods were successful to some extent. Often the product would be high in carbon or low in carbon resulting in either M2C3 phases or MO2 phases respectively. These shortfalls are expected to be overcome by either reducing the carbon content with hydrogen treatment or controlling the M:C ratio in the precursor.
These methods have demonstrated proof-of-concept and now require further development to show that the materials formed can be pressed into pellets or the size of microspheres can be adequately controlled with minimum surface cracking.

1.3.4.2.5 Carbide fuel clad interactions
One aspect of high temperature gas cooled reactor technology is to have a fuel clad system that can withstand the high temperatures required. The most promising systems involve the use of silicon carbide/silicon carbide fibre bonded clad material; however, this may not retain all of the fission products and so a liner may be used. One material that is being considered as a liner is titanium carbide and extensive work was performed to establish the chemical compatibility of uranium carbide and titanium carbide at high temperatures (1800 to 2000°C). Uranium carbide beads of stoichiometry UC1.02 were contacted with a titanium carbide plate and subjected to temperatures of 1800 and 2000°C in a graphite Knudsen cell in order to simulate accident conditions within a nuclear reactor. The objective of this experiment was to assess whether TiC is a suitable barrier material for gas cooled fast reactor fuel. The results show that no visible chemical interaction occurred and the UC bead became detached from the substrate again suggesting that there was no fusion between the materials. Ceramography and microscopy of sectioned parts of the test materials (e.g. Figure 14) support this conclusion. This interesting result suggests that TiC may be suitable as a barrier that is inert with respect to reaction with UC.

1.3.4.3 WP4.2 Reprocessing of carbide fuel
On the general topic of reprocessing, many countries have decided to adopt the once through fuel cycle for either economic considerations or for fear of proliferation of nuclear materials. Only a few countries such as France, Japan, India, the Russian Federation and the United Kingdom have continued to reprocess spent fuel. However, there is currently widespread interest in the world for adopting closed fuel cycles. If fast reactor technology is adopted then the strategy of a closed fuel cycle with multiple recycling of fissile material is absolutely essential; spent nuclear fuel becomes a raw material for reuse of the bred fissile content and the remaining fissile content after irradiation remains too high to be treated as a waste. Contrary to the current situation with thermal reactor systems, where a choice between the once through and closed cycle may be justified, the fast reactor (FR) concept is not sustainable without a closed fuel cycle.
There are two main options to be considered for aqueous reprocessing of spent carbide fuel: oxidative pre-treatment to an oxide followed by dissolution of the resulting oxide or direct dissolution of the carbide Both methods have their pros and cons and both aim to produce an appropriate aqueous solution feed to a solvent extraction process. An additional option of using molten salts to oxidise the carbide and possibly use the resultant melt in a molten salts electrochemical reprocessing flowsheet, has also been considered.
1.3.4.3.1 Pre-oxidation of carbide fuel
The main benefit for the pre-oxidation of carbides is to remove the carbon from the system so that no soluble organics can be produced in the aqueous feed to the solvent extraction process. This strategy has been adopted elsewhere although there were some areas that required further examination:
· What are the safe handling requirements for carbide powders that might be present during fuel chopping and in fuel manufacture?
· Can carbide oxidation using CO2 result in better control of the exothermic reaction that occurs in air and provide a solid that will dissolve in nitric acid, avoiding the formation of insoluble plutonium rich phases?
· What level of volatile fission products are released during high temperature pre-treatments with CO2?
Experiments were performed to answer these questions supported by computer modelling of the oxidation processes.
1.3.4.3.2 Pyrophoricity of uranium carbide
Studies were performed on milled high surface area carbide material to assess the ignition process and establish how the sample size and oxygen concentration influenced the reaction rate and temperature at which a runaway ignition event occurred. Experiments performed in air in a Pyro furnace, with CCD-camera monitoring temperature and thermocouples in direct contact with the 2mm powder bed (surface area 2m2/g) showed an ignition temperature around 100°C, with good reproducibility. The different ignition steps were observed with glowing of the powder, burning, volume increases and sparks spreading over the sample at 340°C.
An atmosphere with lower oxygen content than air (3 % O2 in nitrogen) will allow safe fabrication and handling conditions for future fuel. Ignition has not been observed under these conditions with a mass of 1.5 g. The amount of UC powder and its geometry are important parameters as far as ignition temperature is concerned, as well as the surface area to volume ratio. Very low powder thickness leads to ignition near 200°C, whereas larger amounts of powder lower this ignition temperature to a value around 100°C and then stabilises with higher thickness (2 to 6 mm, or 1.5 to 3 g). CO2 gas emission near 320°C and mass loss were observed using mass spectrometry coupled-TGA.
Furthermore, a small 5-6 mm diameter UC ingot did not ignite in flowing air below 500°C, whereas, for the same heating conditions, the UC powder obtained when grinding the ingot showed pyrophoric behaviour near 100°C. A dense layer of U3O8 was identified in ceramography samples on the surface of the ingot suggesting that the oxide can form a diffusion barrier to oxidation.
The UC oxidation mechanism seems to involve the following:
1) Exothermic reaction of oxygen with uranium carbide: uranium oxidises and carbon
precipitates.
2) Temperature increase of the powder and acceleration of the oxidation kinetics.
3) Ignition with burning and reaction goes on depending on the crucible dimension.
4) Exothermic reaction of residual carbon with oxygen leading to CO2 emission and UC
powder sparks.
5) Final transformation into higher oxides until U3O8.
For safe handling of milled uranium carbide materials, the main considerations are the use of an inert gas to store the powder (< 3 % O2) and to avoid heating the material close to 100°C. Moreover, powders should be stored as much as possible inside small diameter and tall containers, so that the surface exchanging with oxygen remains small and oxygen diffusion inside the powder bed is limited if an incident occurs.
These experimental studies were supported by computer modelling of the oxidation process, see Figure 15, which confirmed the importance of the spalling effect of oxide from carbide pellets as the UC turned to UO2 and then U3O7, U4O9 and U3O8 resulting in a decrease in density, increase in volume and mechanical breakaway from the pellet surface.
1.3.4.3.3 Pre-oxidation of (U,Pu)C microspheres
If an oxidative pre-treatment step is to be applied to carbide fuels then care needs to be taken so that insoluble plutonium rich phases are not generated. One method that shows promise is to oxidise carbide fuels using CO2.
Oxidative treatment of fabricated carbide fuel (U0.8Pu0.2)C was performed at 1000°C in CO2 atmosphere. Initial results suggest that no PuO2 phase separation occurs. For the oxidative treatment of mixed carbide, a TGA oven was used. TGA is advantageous as it allows monitoring of the weight gain/loss within the time on controlled temperature and redox conditions. The gain of the mass was determined to be between 7 and 8 wt.%.
Thermodynamic calculations of (U0.8Pu0.2)C oxidation predict the oxidation product to be a mixture of oxides MO2 (M = U, Pu) and Pu2O3.
Dissolution tests on oxidised and non-oxidised (U0.8Pu0.2)C revealed no significant effect of nitric acid concentration on dissolution kinetics. The amount of organic carbon formed during dissolution is dependent on the solid to solution ratio. Organic carbon analysis could not be performed due to the high alpha activity; however, it was clear that organics did form because of the dark discolouration of the liquor at a 40gHM/L concentration, see Figure 16. This colour is very similar to that obtained with UC dissolved in nitric acid.
These dissolution trials provided qualitative evidence to suggest that pre-oxidation of (U,Pu)C material results in an oxide that is visibly soluble. Also, (U,Pu)C when dissolved directly in nitric acid forms coloured organic solutions similar to UC dissolution in nitric acid.
One potential difficulty with carbide oxidation using CO2 lies in the 14C waste treatment strategy of the reprocessing plant. If CO2 was abated (as in the Sellafield THORP plant) then the amounts of secondary waste generated could be prohibitive and may rule out this reprocessing approach.
1.3.4.3.4 Assessment of volatile fission products
With the success of pre-oxidation demonstrated it was important to establish what the effect of heating to such high temperatures in CO2 atmospheres would be on the volatility of fission products. An InVap system was used to analyse small fragments of irradiated (U,Pu)C fuel. The material was heated to 1000°C in CO2 and the fission product release compared to heating samples that had not been oxidised. Table 3 shows the main fission product results demonstrating which ones gave significant fission product release.
1.3.4.3.5 Direct dissolution of carbides
There are a number of options for the reprocessing of carbide fuel, one of which is the direct dissolution of the fuel using similar technology to proven operating reprocessing plants in the UK and France.
While this option is attractive from a technology readiness perspective, there are a number of technical issues that need to be resolved. The main issue is the production of organic species that remain in the liquor once dissolution is complete. These organic species can interfere with the downstream solvent extraction process used to separate the uranium and plutonium from the fuel. Experiments performed within the ASGARD project assess the kinetics of dissolution, speciation of the aqueous soluble organics and assessed methods of destroying the organics. The resulting solutions were tested for the solvent extraction of uranium and plutonium and compared to similar reference solutions that had no carbon species in them.
Specifically, unirradiated uranium carbide ingots (1g) and blanket fuel pellets (70g), used in the UK fast reactor programme, see Figure 17, have been dissolved under a range of nitric acid concentrations and temperatures.
The amount of nitric acid consumed per mole of uranium carbide and the level of carbon left in solution is established. The importance of the role of nitrous acid has been established and the overall activation energy calculated. One important improvement on the understanding of the dissolution mechanism is the identification of the previously uncharacterised mixture of aromatic organic species. A mixture of polycarboxylic acids with varying amounts of nitration have been positively identified by a range of analytical techniques in addition to the already known oxalic and mellitic acids, see Figure 18.
Precipitates during UC dissolution were also identified as uranyl oxalate. With the solubility of plutonium oxalate much lower than uranyl oxalate then these solids can potentially represent a criticality hazard that should be avoided.
The amount of organic species in solution could not be significantly reduced by altering the dissolver conditions and > 50 % remains in solution once the uranium carbide pellet is dissolved. These organics interfere with the solvent extraction of uranium and plutonium and need to be destroyed to a sufficiently low level such that the Pu losses to the HA waste stream are minimised. Using several techniques, they were characterised as carboxylic acids, mellitic acid, benzenepenta carboxylic acid and unfunctionalised polyaromatics compounds possessing about 3-4 aromatic rings, which tend to be highly functionalised. The use of a boron doped diamond electrode to destroy the remaining organics was shown to be successful when combined with a silver(II) catalyst. The level of organics was reduced to below 0.66 % of the initial carbon in the pellet.
Solvent extraction experiments were used to confirm the success of organics destruction. Solutions containing uranium and plutonium were extracted using TBP/OK in the presence of no organics, organics without destruction and organics with destruction. The solution containing organics did result in higher than acceptable levels of uranium and plutonium remaining in the aqueous phase after 5 contacts with solvent. Conversely, the solution where the organics had been destroyed performed no worse than the solution that had no organics in it. Both uranium and plutonium were backwashed (stripped) from the solvent with no problems. This demonstrates that the organics destruction process was indeed successful.
A knowledge gap on the behaviour of iodine chemistry during carbide dissolution was identified and experiments were performed to understand its potential to form volatile organo-iodine compounds. Reacting iodides with some of the identified organics (oxalic acid, mellitic acid and graphite) under dissolver conditions did not result in detectable amounts of volatile organo-iodine compounds but these were formed when the reactions were performed under irradiation. Further studies are required to quantify this effect.
Dissolution studies on other components of potential gas-cooled fast reactor (GCFR) fuel components have also been performed. Titanium carbide is a potential liner for GCFR fuel and it is likely to be fed to the dissolver with the fuel. Titanium carbide powders were shown to dissolve readily under dissolution conditions and form titanium nitrate plus organic species that can cause the titanium to precipitate from solution at high acidities (≈12 mol/L). This solid is thought to contain oxalate and nitrate groups. Further studies are required on titanium carbide ceramics.
The titanium carbide liner is designed to keep the fission product gases from diffusing through a silicon carbide/silicon carbide fibre fuel clad. Dissolution studies on real samples have demonstrated that no detectable silicon carbide dissolves under dissolver conditions.
1.3.4.4 WP4.3 Carbide processing with molten salts
Within this part of the project an alternative view to hydrometallurgy methods were explored. The overall objectives were to establish whether molten salts could be used to react/oxidise carbides and provide a molten salt solution that could then be integrated within a pyro- reprocessing flowsheet to recover uranium and plutonium.
1.3.4.4.1 Chemical pre-treatment of carbide fuels in molten salts
A pellet of uranium carbide fuel can be thermally treated for 30 min at 800oC in molten CaCl2-NaCl (46 mol % CaCl2) in the presence of 10 wt % CaO under argon to give a uranium (III) chloride solution.
Three types of TRISO simulated fuel particles incorporating alumina kernel have been thermally treated for 3h at 1000oC in molten chlorides. The resulting samples have been characterised by SEM/EDX and stripped coated layers of TRISO particles have been observed, confirming that the process of TRISO oxidation degradation has taken place under applied thermal treatment conditions.
As a result of thermal treatment, the β-SiC based cladding (SiC/Ta/SiC) has dissolved in molten CaCl2 containing 12.5 wt % of CaO at 1000oC for 12 h, thus demonstrating that the molten salts treatment method works efficiently for the SiC based fuel cladding materials. Electrolytic dissolution of UC giving U(III) in molten LiCl-KCl eutectic has been proven to occur, during which black particulates float to the melt surface that are most likely insoluble carbon based materials.

Potential Impact:
1.4 Description of the potential impact, the main dissemination activities and the exploitation of results
The impact of the ASGARD project is multi-fold both in achievements, human resources, spatial effects and time.
It is naturally difficult to rank these different impacts in some order of importance but considering the human resource situation in Europe today with a lack of younger specialists the human resource impact is vital. By the extensive education and training program of ASGARD many younger scientists got the possibility, not only to advance their knowledge in a narrow field of science as shown by their PhD theses but they also got a much wider scope of both basic education and also hands on training on real material. This latter kind of education is very rare and those with the special knowledge about safe and secure handling of substantial amounts of radioactive material are highly sought after and need for a continuous safe operation of the nuclear facilities in Europe. Thus, by the ASGARD project a number of skilled and trained young scientists have been made available for the European research and industrial communities.
By the nature of the project the main tangible impact is the scientific results significantly increasing the knowledge of the different fuels handled in the project. This will create a good knowledge base for future studies in the fuels as well as advancing the scientific boundaries in some key areas of science on a deep basic level. Examples of such advances are the identification of new mixes species in molybdenum based systems as well as basic chemical behaviour of plutonium.
There were also significant achievements increasing the industrial applicability of advanced fuel fabrication. Naturally this is an important impact if there is to be any advancement from the rather poor behaviour of the oxide type of nuclear fuels to the more advanced types. This development is strengthened by the fact that companies in both the USA and the Russian Federation are currently developing e.g. nitrides for use in current light water reactors. Thus by the progress in the ASGARD project Europe has a possibility, if selected, to be on the fore front of developing a sustainable nuclear fuel cycle based on the desirable properties of the advanced fuels investigated.
A tangible result in this direction developed within ASGARD is the microwave gelation process at PSI. By using this technique the previously industrially shunned sol-gel technique will be considerably easier to implement. In turn this will also enhance the possibility of producing new fuels directly from the solutions resulting from a separation/recycling process.
Taking these achievements together it is easy to see a picture where the results of the ASGARD project with respect to manufacture and dissolution of advanced fuels together with the increased industrial applicability of the solution based sol-gel fabrication route clearly has an impact on potential future sustainable nuclear fuel cycles.
Through not only ASGARD but taken into a greater context society has now an increased ability to benefit from the extended efficiency of the advanced fuel utilisation but it will also be possible to create a more circular nuclear economy saving both natural resources as well as energy in the future.
It is clear that the ASGARD project was a success when relating the planned performance indicators as defined in the original project plan to those really obtained as presented in Table 4. All quantitative targets have been achieved with some of the target values significantly surpassed.
Dissemination and exploitation
The dissemination of the knowledge generated in a project like ASGARD is divided into several parts. Naturally for a scientific project the main channel for spreading the knowledge is through scientific publications. By the pure nature of the ASGAD project and the fact that it is the first of its kind most of the publications are published rather late. Thus at the writing of this report 17 scientific papers were published in reviewed journals and another 10 were either submitted or already accepted. It is expected that this number will be about doubled when the total outcome of the project is published.
Another important channel for spreading the knowledge generated is through conference presentations. In this context the ASGARD project has been working two fold. Both through the support generated by the mobility plan, as discussed above, but also through active promotion of the younger participants to present their results by themselves at different relevant scientific events. The total number of such conference presentations though the duration of ASGARD is 61.
In order to increase the visibility of the ASGARD project as well as to gain important external input to the work performed three international events were organised. Originally some of these were planned to be stand-alone events but it became clear that the wealth of scientific meetings and conferences was significant why it was deemed more efficient to join ASGARD’s sessions to already existing, high quality conferences. The conferences thus including dedicated ASGARD sessions were: ATALANTE (2012), RADCHEM (2014) and TOPFUELS (2015). In the latter case the session was co organised by the PELGRIMM project. In these sessions which typically were between 0.5- 1 day the participation between project presentations and other external, relevant presentations was about 50/50. This division ensure that both the objectives of ASGARD visibility as well as important external input were met in a satisfactory manner.
The project generated three patent applications:
• CEA: Method for colloidal preparation of a metal carbide, said metal carbide thus prepared and uses thereof – WO2016020496
• ICHTJ: Method for preparing of uranium-neodymium dioxide spherical grains Double Extraction Process – PL403817
• ICHTJ: A method for producing uranium carbide of spherical and irregular grains as a precursor to fuel for reactors of the new, fourth generation – PL414768
Additionally, 31 Exploitable Foregrounds have been identified and described by ASGARD partners and reported through the D1.3.7 Exploitation Plan including their plans for further use in various domains (further research, education and training, dissemination and other purposes). The most of the results are only protected under IPR arrangements of the ASGARD Consortium Agreement.
The project website was launched in the beginning of the project at www.asgardproject.eu and updated throughout the project. It attracted more than 9 000 visitors. There were 52 news published since the webpage launch, 23 downloadable project publications including public deliverables, generic project information materials (leaflet, logo, presentation, poster etc.) and promotional materials related to the project events. The project webpage will be continued after the project end.

List of Websites:
www.asgardproject.eu
Prof Christian Ekberg
Nuclear Chemistry
Department of Chemical and Biological Engineering
Chalmers University of Technology
S-412 96 Göteborg
Sweden

Tel: +46-31-7722801
Fax: +46-31-7722931

email: info@asgardproject.eu

Related information

Contact

Karin Ryrberg, (Head of Administration)
Tel.: +46 31 7723210
E-mail

Subjects

Nuclear Fission
Record Number: 191937 / Last updated on: 2016-11-16