Elsevier

Carbon

Volume 86, May 2015, Pages 174-187
Carbon

Milled graphite as a pertinent analogue of French UNGG reactor graphite waste for a CO2 gasification-based treatment

https://doi.org/10.1016/j.carbon.2014.12.060Get rights and content

Abstract

Graphite has been used in gas-cooled nuclear reactors as a neutron moderator. The dismantling of nuclear reactors will generate significant amounts of graphite waste. Neutron irradiation is responsible for 14C formation in graphite, and it leads to severe structural and nanostructural degradations. At high neutron fluence, nanoporous turbostratic carbon is formed from original lamellar graphite. This phase is supposed to be especially 14C enriched. An original 14C extraction process was proposed: to “decontaminate” graphite waste from 14C by selectively gasifying such degraded areas, without entirely consuming the graphite waste. To specify the operating conditions, milled graphite was used as a nonradioactive analogue. Raman microspectrometry and transmission electron microscopy techniques show that neutron irradiation and milling lead to similar multiscale organization, and especially nanoporous carbon formation. Thermogravimetry experiments were then carried out between 800 and 1100 °C, at a CO2 pressure of 0.1 MPa. To determine the best temperature range allowing a nanoporous component selective gasification, Raman microspectrometry analysis was coupled with transmission electron microscopy observations on the residues obtained for each gasification temperature. The 950–1000 °C temperature range is the most efficient allowing a complete elimination of degraded areas supposed to be representative of nuclear graphite waste 14C-rich areas.

Introduction

Graphite has been used in French gas-cooled nuclear reactors, called Uranium Naturel Graphite Gaz (UNGG) reactors, as a neutron moderator. In France, these reactors have no longer been operated since 1994. Their dismantling will generate significant amounts of nuclear waste including 23,000 tons of neutron-irradiated graphite. At the world scale, the amount of graphite waste is estimated to be 250,000 tons. Neutron irradiation is responsible for the formation of some radionuclides such as 36Cl, 14C, and H3 [1], [2], [3]. As a consequence, graphite waste is classified in France as low-level long-lived waste. 14C is a major contributor to the activity of this nuclear waste. Moreover, it will be the main contributor after a thousand years of radioactive decay. Several long-term management options are being considered in France for graphite waste [1], [3], including disposal with or without a partial decontamination pretreatment [4], [5], [6].

Neutron irradiation induces displacement cascades in the graphite matrix and 14C creation. Structural degradation is commonly quantified by using X-ray diffraction (XRD) [7], [8], Raman microspectrometry, scanning electron microscopy (SEM), and transmission electron microscopy (TEM) [5], [6]. Thanks to XRD and Raman microspectrometry, structural degradations are evidenced: the graphite tri-periodic order is lost, the crystallite size strongly decreases, and, in the most degraded areas, such irradiated carbons become entirely turbostratic. SEM and TEM show dramatic morphological and textural changes: planar graphite lamellae become more and more crumpled with increasing neutron fluence. High-resolution transmission electron microscopy (HRTEM) allows direct imaging of the structural and nanostructural irradiation damages. In the most irradiated parts, the random orientation of the nanometer-sized stacks of graphene layers is responsible for the occurrence of a nanoporous turbostratic carbon. Such disorder is very similar to that induced by ion irradiation [9]. It is thus possible that a part of 14C is preferentially concentrated in such nanoporous carbon graphite waste. It is the basic hypothesis of our method to extract a part of 14C from the graphite waste matrix [4]. Note that the remaining 14C in this graphite waste was produced from 13C initially linked to the matrix. Indeed, 14C produced from 14N was probably released during reactor operation [3]. That is why, in the French UNGG reactor graphite waste case, this 14C is hardly extractible by thermal treatment [10].

Originally, milled graphite was synthesized to improve the Li-ion battery using graphite cathodes in the place of perfect graphite [11]. However, it is remarkable that very similar structural and nanostructural changes were observed by HRTEM and Raman microspectrometry on this milled graphite in comparison with neutron-irradiated graphite. Even if the mechanisms of such damage creation are obviously different, these structural and nanostructural similarities were exploited here in order to have a pertinent nonradioactive analogue of nuclear graphite waste. It will be especially used to specify the effect of the CO2 gasification process on the selective elimination of the most degraded areas from irradiated nuclear graphite waste.

Very disordered areas of irradiated graphite are the result of very strong neutron fluence and they probably should contain 14C, whereas less damaged graphite may remain 14C less rich. Thus, our basic hypothesis is that 14C concentration is coupled with the nanostructure, the highest 14C concentration corresponding to the most nanoporous carbons. Moreover, TEM studies show that neutron irradiation of graphite leads to systematically heterogeneous carbon at the nanostructure scale, that is, between micrometer scale and nanometer scale, regardless of the global conditions of neutron fluence and operation temperature [4], [5], [6].

Besides, it is well known that carbon reactivity with carbon dioxide strongly depends on the carbon nanostructure (previously called “microtexture”) [12], [13]. Indeed, as the active sites are mainly localized on the edge of graphene sheets [14], [15], their concentrations must be maximum in the most nanoporous areas. The effective reactivity, that is, the relative weight loss, of perfect lamellar graphite is very low, while it is very high for the most disordered carbons, at the same gasification conditions (gas pressure and temperature).

Those are the reasons why Rouzaud et al. proposed at the Annual World Conference on Carbon 2011 a process to “decontaminate” graphite waste by selectively gasifying this nanoporous carbon with carbon dioxide. Thus, 14C activity of graphite waste could be reduced with a weak weight loss [4]. Furthermore, numerous works carried out on coke gasification in blast furnaces incited us to select carbon dioxide as the best selective oxidant gas. Oxygen entirely gasifies carbon without selectivity according to the different carbon nanostructures. Steam is avoided due to possible hydrogen and tar formation. Moreover, the produced CO2 cannot be posttreated as easily as CO.

Thus, the CO2 gasification process could be a solution to reduce the 14C activity of graphite waste. We are examining optimal conditions for this process allowing a complete elimination of the degraded areas, that is, the supposed 14C-rich areas. At the same time, the aim is to avoid gasifying the lamellar and 14C-poor component of graphite waste so as to produce as less as possible 12CO (low weight loss) and thus increase the 14C-selective extraction.

A mechanism was proposed for carbon gasification with CO2 based on the Boudouard equilibrium [16], [17], [18], [19]:C(s)+CO2(g)2CO(g)

The first step of the chemical reaction is considered to be the CO2 chemisorption on active sites of C(s) responsible for the C(O) complex formation with a kinetic constant of i1. In the chemically controlled regime of reaction, this chemisorption step depends on the nature of the active sites [20], [21], [22]. For the C–CO2 reaction, actives sites are known to be on the edges of graphene layers [14], [15], and active-site density is also an important parameter [12]. However, CO formation is known to inhibit the reaction [16], [17], [23] with the kinetic constant of inhibition reaction i2:C(s)+CO2(g)C(O)+CO(g)

The second and rate-determinant step is the desorption of CO from the C(O) complex with a kinetic constant i3:C(O)CO(g)

The Langmuir–Hinshelwood [17] model is usually used to define the gasification rate as a function of kinetic constants (k1 = i1; k2 = i2/i3; k3 = i1/i3) and partial pressures of CO2 and CO:r=k1pCOz1+k2pCO+k3pCOz

This model is available in the chemical regime of the reaction [17]. Indeed, in the internal or external diffusion regime, carbon gasification is also monitored by the mass transfers. There is a transition temperature range between chemical and porous diffusion regimes [17], [18], [19]. This temperature range depends on many parameters such as CO2 pressure, particle size [18], and particularly on the carbon nanostructure [12], [24]. For instance, the graphite gasification regime at 1000 °C and CO2 atmospheric pressure is known to be chemically controlled, whereas it is a porous diffusion regime for cokes [24]. Therefore, active-site density and accessibility are also known to be key parameters. They obviously depend on the carbon nanostructure, which governs the pore size distribution and the number of accessible edge layers.

Experiments with radioactive waste require many cautions and time. By contrast, nonradioactive analogue materials allow a quicker and easier parametric study of the foreseen CO2 gasification process. Based on the characterization of French UNGG graphite waste, milled graphite was synthesized in order to reproduce neutron irradiation structural and nanostructural damages and to relevantly test the decontamination process. Our aim was to determine the best CO2 gasification temperature, at a CO2 pressure of 0.1 MPa and for two different gasification times to selectively eliminate the nanoporous component. This information is required in order to apply a selective elimination of the most damaged areas of nuclear waste with the weakest global weight loss as possible.

Section snippets

Milled graphite synthesis as a pertinent analogue of French UNGG graphite waste

Previous studies have shown that the fraction of nanoporous carbon in the milled graphite can be monitored by controlling the milling duration and the atmosphere composition in the milling jar [25], [26]. In the present work, the precursor is the Timcall SLX-50 graphite, with a particle size of 50 μm. The structure and nanostructure are very close between the non-milled graphite and the virgin nuclear graphite. To reproduce well the most damaged nuclear graphite areas, six milling cycles of 3 h,

Milled graphite multiscale characterization

First of all, we verified that our milled graphite is a relevant structural and nanostructural analogue of the neutron-irradiated graphite. For that, characterizations were carried out on both neutron-irradiated graphite from French UNGG reactors and milled graphite. Raman spectra were first recorded, and then HRTEM confirmed that structural and nanostructural characteristics, including heterogeneities, were similar to that of neutron-irradiated graphite.

At the micrometer scale, the Raman

Selectivity, gasification temperatures, and diffusion regimes

Regardless of the gasification duration studied, the degraded areas of milled graphite, characterized by Raman ratios ID1/IG > 0.5, are selectively gasified at 950 and 1000 °C. The nanoporous components of microparticles completely disappear in the gasification residues. As shown in this study, these gasification conditions well allow a complete and selective gasification of nanoporous components, whereas lamellar ones are preserved.

In the case of heterogeneous materials made of an association of

Conclusion

As recently proposed by our team, CO2 gasification could be an original way to extract a significant part of 14C from neutron-irradiated graphite waste. To test this process in nonradioactive conditions, specific milled graphite analogues were synthesized. As observed for neutron-irradiated graphite from French UNGG reactors, milled graphite multiscale organization is severely damaged. Despite their obviously different causes, TEM images show these degradations to be similar: they occur

Acknowledgments

This work is a part of a thesis financially supported by Andra with the partnership of CEA, CNRS, ENS, and EDF. The authors warmly thank Lionel Sejourne, Jérôme Comte, Catherine Desserouer, Laurence Petit, Laurent Petit, and Stéphane Catherin.

References (32)

Cited by (17)

  • Utilization of biomass ash for upgrading petroleum coke gasification: Effect of soluble and insoluble components

    2020, Energy
    Citation Excerpt :

    Compared with washed ash (Ca-rich), the ash leachate (K-rich) was more likely to cause the reaction control change with the gasification proceeding. Gasification reactivity strongly depended on the carbon nanostructure [41]. By analyzing the carbon structure evolution, it could be possible to gain some insights about the catalytic role of different ash-based catalysts.

View all citing articles on Scopus
View full text