Influence of gamma ray irradiation on metakaolin based sodium geopolymer

Journal of Nuclear Materials 443 (2013) 311–315

Contents lists available at ScienceDirect

Journal of Nuclear Materials journal homepage: www.elsevier.com/locate/jnucmat

Influence of gamma ray irradiation on metakaolin based sodium geopolymer D. Lambertin a,⇑, C. Boher a, A. Dannoux-Papin b, K. Galliez a, A. Rooses a, F. Frizon a a b

CEA, DEN, DTCD/SPDE/LP2C, F-30207 Bagnols-sur-Cèze, France CEA, DEN, DTCD/SPDE/LCFI, F-30207 Bagnols-sur-Cèze, France

a r t i c l e

i n f o

Article history: Received 20 March 2013 Accepted 25 June 2013 Available online 4 July 2013

a b s t r a c t Effects of gamma irradiation on metakaolin based Na-geopolymer have been investigated by external irradiation. The experiments were carried out in a gamma irradiator with 60Co sources up to 1000 kGy. Various Na-geopolymer with three H2O/Na2O ratios have been studied in terms of hydrogen radiolytic yield. The results show that hydrogen production increases linearly with water content. Gamma irradiation effects on Na-geopolymer microstructure have been investigated with porosity measurements and X-ray pair distribution function analysis. A change of pore size distribution and a structural relaxation have been found after gamma ray irradiation. Ó 2013 Elsevier B.V. All rights reserved.

1. Introduction Geopolymers are a class of inorganic binders which are solid, aluminosilicate, charge-balanced by alkali cations such as Na+, K+ or Cs+ [1]. They are generally formed by reacting aluminosilicate powder (metakaolin, fly-ash) with highly caustic, aqueous alkali silicate solution. Geopolymers have been studied for immobilization of low and intermediate level nuclear waste [2] especially for Cs and Sr encapsulation [3–5] or Mg–Zr alloy fuel cladding immobilization [6]. Moreover, geopolymer is already used in nuclear industry to encapsulate sludges and organic ion exchange resins in Slovakia [7]. As cement matrices, geopolymers contain water in their pores [8] that could release hydrogen, the main product gas of water radiolysis. Previous experiments demonstrate that the radiolytic yield of hydrogen production G0(H2) determined at 50 kGy on a Na-geopolymer dried under ambient atmosphere during one month after synthesis is 0.061  107 mol/J [9]. This result indicates a lower H2 production by water radiolysis than that obtained in cement matrices [10,11]. In the present work, the effects of gamma irradiation on air-tightly stored sodium geopolymer material have been studied in terms of radiolytic hydrogen yields, mechanical properties, porosity measurements and structural characteristics with X-ray pair distribution function. 2. Experimental The sodium geopolymer samples Al2O33.6SiO2Na2OxH2O (x = 11–13) thereafter called Na-Geo – have been prepared by ⇑ Corresponding author. Tel.: +33 466791480; fax: +33 466397871. E-mail address: [email protected] (D. Lambertin). 0022-3115/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jnucmat.2013.06.044

mixing metakaolin powder with sodium silicate solutions. Silicate solutions 1.2SiO2Na2OxH2O were prepared by dissolving amorphous silica (TIXOSIL 38, Rhodia factory) in sodium hydroxide solutions (NaOH pellets, Prolabo, 98%). The composition of metakaolin (2.4SiO2Al2O3), Pieri PREMIX MK from Grace Constructions, is given in Table 1. After demoulding, the samples have been stored in sealed bags in order to prevent water evaporation. The water content of Na-Geo Al2O33.6SiO2Na2OxH2O with different water composition has been determined by water resaturation method (EN 1936) and is reported in Table 2. To quantify the hydrogen gas produced by gamma irradiation, about 20 g of geopolymer samples were irradiated in 200 cm3 sealed glass containers in an industrial irradiator using 60Co csource (IONISOS – Dagneux, France) at a dose rate of 600 Gy/h until 750 kGy under argon atmosphere at room temperature. Dosimetry was realised with Perspex devices. After irradiation, the atmosphere of the glass container is analysed with a gas chromatography equipped with a TCD detector (Varian, CP3800). The standard error on the calculated radiolytic yield is about 10%. Compressive strength measurements were performed on 4  4  16 cm prismatic samples (EN196-1). The samples were obtained by pouring fresh Na-geopolymer paste into 4  4  16 cm PTFE moulds, vibrated for a few seconds, and sealed from the atmosphere. Samples were cured for 4 days at 20 °C and at atmospheric pressure before removal from the mould and stored in sealed bags. The pore microstructure was studied with isothermal nitrogen adsorption–desorption measurement performed with a Micromeritics ASAP 2020 device and with Mercury Intrusion Porosimeter (MIP) from Micromeritics Autopore IV 9500 with a maximum injection pressure of 420 MPa. Before the porosity measurements,

312

D. Lambertin et al. / Journal of Nuclear Materials 443 (2013) 311–315

3. Results and discussion

Table 1 Chemical composition of metakaolin.

3.1. Hydrogen radiolytic yield G(H2)

Oxide weight composition (%)

Metakaolin

CaO

SiO2

Al2O3

Fe2O3

K2O

TiO2

0.1

54.4

38.4

1.3

0.6

1.6

Table 2 Hydrogen production radiolytic yields of Na-Geo samples with different water contents obtained at 750 kGy. Sample

Storage

Water content (wt%)

G(H2) (107 mol/ J)

Reference

Al2O33.6SiO2Na2O11H2O Al2O33.6SiO2Na2O12H2O Al2O33.6SiO2Na2O13H2O Al2O33.6SiO2Na2O11H2O Portland cement (water/ cement = 0.4) Water in free volume

Sealed Sealed Sealed Air Sealed

32.5 34.1 35.9 18.1 28.6

0.090 0.102 0.113 0.061 0.80

This work This work This work [9] [22]

–

100

0.45

[11]

freeze-drying technique was used as a low alteration technique of the pore structure [12]. Phase analysis was made with a Siemens (Model D5005) X-ray powder diffractometer equipped with a Cu source (kKa = 1.5418 Å, 2h range: 5–70°, steps of 0.02°, acquisition time of 1 h). Structural characterisation of unheated geopolymers has been limited by their amorphous structure [2]. X-ray pair distribution function analysis (PDF) is appropriate for the study of materials with short coherence lengths including geopolymer materials [13–18]. PDF analysis, on the other hand, concerns the entire signal including Bragg peaks and diffuse scattering. The pair distribution function (PDF) represents the distribution of interatomic distances in a compound, regardless of its crystalline state, determined experimentally by a Fourier transform of the powder pattern. The experimental PDF is given by the following equation [19]:

GðrÞ ¼ 4pr½qðrÞ  q0  ¼

2

p

Z

Q max

Q ½SðQ Þ  1 sinðQrÞdQ

ð1Þ

Q min

Geopolymers are porous materials. Like in cement matrices, water is contained in the pores of the material after setting. In literature, water radiolysis is well known to engage a series of reactions called Allen cycle with the formation of eight primary species    (4 molecules: H2, H2O2, OH, H3O+ and 4 radicals: e aq , H , OH , HO2 ) [10]. One of the major stable products of water radiolysis is hydrogen gas. In free volume, the primary G(H2) issued from water radiolysis is 0.45  107 mol/J [11]. In a hydraulic binder, several parameters could interfere on the H2 production such as the global water content, the pore size distribution or the interstitial water chemical composition. However, these effects have not been studied in geopolymers matrices to our knowledge. The results on NaGeo samples Al2O33.6SiO2Na2OxH2O are shown in Table 2. From 32.5 to 35.9 wt% of water in the Na-Geo, the radiolytic yield G(H2) were between 0.090  107 and 0.113  107 mol/J. First, data indicate that the hydrogen production under gamma radiation increases with the water content in Na-Geo. The comparison of these results to the hydrogen radiolytic yield of 0.061  107 mol/J obtained in the previous study [9] indicate that in our case, hydrogen production was larger. Indeed, as the Na-Geo sample in the former study has been stored under air before irradiation, the water content was reduced to about 18% due to evaporation. As the water content also affects the pore size of the Na-Geo, an influence of the pore size on the hydrogen production might be expected but cannot be estimated through these few tests. On the other hand, these radiolytic yield G(H2) are lower than the G(H2) of 0.8  107 mol/J obtained in Portland cement matrices with a water/cement ratio of 0.4 (water content = 28.6 wt%) [22]. Even though the water content in the Na-Geo is higher, G(H2) remain much lower than the cement reference value. This result can be explained not only by the difference in pore structure between cements and geopolymers but also by the disparity of pore solution [23,24] with regard to pH and chemical constitution. Indeed evolution of the primary yields of species depends on the alkaline medium, especially in the pH 12–14 range with cement and geopolymer pore solutions. The primary G(H2) can vary from 0.435  107 mol/J to 0.384  107 mol/J between pH 12.45 and 14 [10]. Besides, the ionic constitution (Ca, Na, Al, Si, K, etc.) of the pore solution are very different for cement and geopolymer [23,24]. These differences must significantly modify the stability of intermediates related to the radiolytic recombination of H2 and therefore affect the production of hydrogen gas.

where G(r) is the reduced PDF, q(r) is the microscopic pair density, q0 the numeric density of the compound (number of atoms per unit

3.2. Microstructure

volume), and S(Q) normalised scattering intensity, known as the total static scattering function. It is important to obtain diffraction data with a high scattering vector (Q = 4psinh/k) in order to maximise the resolution after the Fourier transform. We therefore used a PANalitycal (Model X’Pert Powder) X-ray powder diffractometer equipped with a Mo source (kKa = 0.70926 Å, 2h range: 1.5–155.66°, steps of 0.017°, the total acquisition is the average of 4 runs recorded over 4 h) to attain Qmax = 17 Å1. The PDF and standard corrections were calculated with pdfgetX2 [20]. The calculated pair distribution function from structure was obtained with PDFGUI software [21]. Due to a limited Q resolution, a PDF Gaussian dampening (Qdamp) envelope has to be introduced in the refinement. The damping factor Qdamp was assigned to a value of 0.05. Another parameter s-ratio is a reduction factor for PDF peak width accounting for correlated motion of bonded atoms.

Radiations such as gamma ray could induce structural changes or even damages in solid materials. In order to reveal any structural modification on Na-Geo related to gamma radiation effect, compressive strength measurements have been done on non-irradiated and irradiated samples and for different doses at 50, 500 and 1000 kGy (Fig. 1). The results have shown an increase in compressive resistance by 10% after irradiation for the three absorbed doses. These observations assume a densification in the geopolymer network structure under gamma irradiation. To provide explanations to this phenomenon, the porous network has been further characterised with nitrogen adsorption porosimetry measurements on Al2O33.6SiO2Na2O11H2O in nonirradiated and irradiated conditions. Fig. 2 presents the nitrogen sorption/desorption isotherms on non-irradiated and irradiated Na-Geo samples. The isotherms of non-irradiated Na-Geo present at first a pronounced stage of the

D. Lambertin et al. / Journal of Nuclear Materials 443 (2013) 311–315

313

Fig. 1. Compressive strengths of Na-Geo samples without and with gamma irradiation exposure at 50, 500 and 1000 kGy.

monolayer formation of physisorbed N2, then a multilayer formation for higher relative pressures, and a hysteresis loop on the desorption isotherm which leads to identification of the isotherm profiles as Type IV in the BDDT (Brunauer–Deming–Deming–Teller) system. This is typical of mesoporous materials having pore diameters between 2.0 nm and 50.0 nm. The hysteresis loop with parallel and nearly vertical branches is identified as H1 type (IUPAC), which is characteristic of cylindrical pores with high pore connectivity and often reported for materials that consisted of assemblages of rigidly joint particles [25,26]. Close to the saturation vapour pressure, the adsorption isotherms do not clearly level off which is characteristic of a pore structure theoretically identified as ‘‘open end slit geometry’’ [27] which highlights the presence of a high interconnectivity in the mesoporous structure. In the case of irradiated geopolymer, the total absorbed nitrogen quantity is mostly similar to the non-irradiated material, which indicates the total volume for pores between 1 nm and 20 nm (pore-size investigated by N2 sorption/desorption) remains about the same. The slopes on sorption and desorption isotherms for irradiated sample are less vertical than those for irradiated samples. This underlines that the pore size distribution is larger after irradiation. Moreover, the sorption/desorption isotherms obtained on materials after irradiation exhibit two inflection points

Fig. 2. Nitrogen sorption/desorption isotherms of Na-Geo (solid line) and gamma ray irradiated Na-Geo (dotted line) and their respective pore size distributions calculated with the BJH method.

Fig. 3. Pore size distribution of Na-Geo (solid line) and gamma ray irradiated NaGeo (dotted line).

which is characteristic of a pore size distribution change to monomodal to bimodal nature [28]. The pore size has been calculated with the Barrett, Joyner, and Halenda (BJH) method (Fig. 2). However, the resulting pore size range is too close to the upper limit of the BJH method validity. The mercury intrusion porosimetry was therefore more appropriate to investigate the pore size distribution for pores with diameter larger than 20 nm. The pore access diameter distribution is presented in Fig. 3. The non-irradiated material presented a sharp and monomodal pore size distribution centred at 12 nm. As for the irradiated material, the porous distribution is bimodal with a sharp population centred at 7 nm and a broad population from 10 nm to 50 nm. The results coincide with the observations of the isotherms with regard to the enlargement of the pore size distribution and the apparition of two populations after irradiation instead of one for non-irradiated samples.

Fig. 4. X-ray diffractograms of Na-Geo samples before and after exposure to gamma ray irradiation, and heated at 1000 °C (non irradiated).

314

D. Lambertin et al. / Journal of Nuclear Materials 443 (2013) 311–315

3.3. X-ray pair distribution function

Fig. 5. Experimental pair distribution function of Na-Geo (solid line) and gamma ray irradiated Na-Geo (dotted line).

An explanation of the enlargement of the pore size population could be the result of the thin pore wall reorganisation under irradiation, implying the decrease of a certain pore size population in favour of larger pores.

The phase composition of the Na-Geo samples was analysed by XRD before and after exposure to gamma irradiation (Fig. 4). The broad diffuse hump in the Na-Geo pattern indicates the material is predominantly amorphous. The only Bragg peaks detected in this pattern were due to quartz and anatase, that are form unreacted metakaolin powders. After irradiation at 750 kGy, the pattern of Na-Geo does not display any noticeable changes. Consequently, the gamma irradiation does not involve formation of crystalline phases in the geopolymer network. The XRD analysis of Na-Geo fired at 1000 °C in a furnace shows that the broad hump of the amorphous aluminosilicate is replaced by the peaks due to nepheline structure NaAlSiO4 (JCPDS 0350424). This indicates that the chemical configuration of the amorphous aluminosilicates of the Na-Geo cured at room temperature is closely related to the nepheline stoichiometry. The atomic structure of the amorphous Na-Geo has been assessed by the calculation of pair distribution functions (PDFs) from XRD diffractograms through the use of Fourier transform as described in experimental part. Fig. 5 presents the experimental pair distribution function of Na-Geo before and after gamma ray irradiation. The correlation peaks have been assigned to T–O, O–O, T–T and Na–T distances (T = Si or Al in tetrahedral site) after identification with the calculated partial PDF of nepheline from Kahlenberg et al. structure [29] (Fig. 6) by using the PDFGUI software [21] (Qdamp = 0.05 and s-ratio = 1). The T–O, O–O and T–Na peak correlations at 1.6 Å, 2.64 Å and 3.1 Å, respectively, are clearly not affected by the gamma ray irradiation whereas a shift towards low r of the T–O correlation at 4.2 Å is observed. This T–O distance can be assigned to the interaction of atoms from neighbouring tetrahedrons. Previous studies on silica glass material under gamma ray have shown [30,31] that a relaxation process releases some of the energy stored in the structure, attended by a decrease of average bridging bond Si–O–Si angle, leading to a densification. The shift of T–O correlation observed between the non-irradiated and the irradiated samples can be attributed to a relaxation process of the microstructure at the network scale. 4. Conclusion  H2 radiolytic yield of Na-geopolymer is included between 0.090  107 and 0.113  107 mol/J and mainly depends on the water content.  A change of Na-geopolymer pore size distribution has been noticed after gamma ray irradiation at 750 kGy  Pair distribution function and compressive strengths revealed structural relaxation after gamma ray irradiation of Nageopolymer

References

Fig. 6. Calculated pair distribution function of Nepheline from Kahlenberg et al. structure [29] on PDFGUI [21] (Qdamp = 0.05 and sratio = 1) with the atom specific weighted partial pair distribution function.

[1] J. Davidovits, J. Therm. Anal. 37 (1991) 1633–1656. [2] E.R. Vance, D.S. Perera, in: J.L. Provis, J.S.J. van Deventer (Eds.), Geopolymers, Structure, Processing, Properties and Industrial Applications, Woodhead Publishing Limited, Cambridge, UK, 2009. [3] M.G. Blackford, J.V. Hanna, K.J. Pike, E.R. Vance, D.S. Perera, J. Am. Ceram. Soc. 90 (2007) 1193–1199. [4] S. Berger, F. Frizon, C. Joussot-Dubien, Adv. Appl. Ceram. 108 (2009) 412–417. [5] J.L. Provis, P.A. Walls, J.S.J. van Deventer, Chem. Eng. Sci. 63 (2008) 4480–4489. [6] D. Lambertin, F. Frizon, F. Bart, Surf. Coat. Technol. 206 (2012) 4567–4573. [7] D. Majersy, S. Sekely, M. Breza, IAEA RCM on Behaviour of Cementitous Materials in Long Term Storage and Disposal of Radioactive Waste Meeting, IAEA, Vienna, Austria, MoscowRussia, 2007. [8] P. Duxson, G.C. Lukey, J.S.J. van Deventer, J. Mater. Sci. 42 (2007) 3044–3054. [9] A. Rooses, P. Steins, A. Dannoux-Papin, D. Lambertin, A. Poulesquen, F. Frizon, Appl. Clay Sci. 73 (2013) 86–92. [10] P. Bouniol, E. Bjergbakke, J. Nucl. Mater. 372 (2008) 1–15.

D. Lambertin et al. / Journal of Nuclear Materials 443 (2013) 311–315 [11] E. Bjergbakke, Input Files for Computer Simulation of Water Radiolysis, Risø National Laboratory, Roskilde, Denmark, 1984. [12] C. Gallé, Cem. Concr. Res. 31 (2001) 1467–1477. [13] J.L. Bell, P. Sarin, P.E. Driemeyer, R.P. Haggerty, P.J. Chupas, W.M. Kriven, J. Mater. Chem. 18 (2008) 5974–5981. [14] J.L. Bell, P. Sarin, J.L. Provis, R.P. Haggerty, P.E. Driemeyer, P.J. Chupas, J.S.J. van Deventer, W.M. Kriven, Chem. Mater. 20 (2008) 4768–4776. [15] C.E. White, J.L. Provis, A. Llobet, T. Proffen, J.S.J. van Deventer, J. Am. Ceram. Soc. 94 (2011) 3532–3539. [16] C.E. White, J.L. Provis, T. Proffen, D.P. Riley, J.S.J. van Deventer, Phys. Chem. Chem. Phys. 12 (2010) 3239–3245. [17] C.E. White, J.L. Provis, T. Proffen, J.S.J. van Deventer, J. Am. Ceram. Soc. 93 (2010) 3486–3492. [18] C. Meral, C.J. Benmore, P.J.M. Monteiro, Cem. Concr. Res. 41 (2011) 696–710. [19] T. Egami, S.J.L. Billinge, Underneath the Bragg peaks: Structural Analysis of Complex Materials, Pergamon Press, 2003. [20] X. Qiu, J.W. Thompson, S.J.L. Billinge, J. Appl. Crystallogr. 37 (2004) 678.

315

[21] C.L. Farrow, P. Juhas, J.W. Liu, D. Bryndin, E.S. Bozin, J. Bloch, T. Proffen, S.J.L. Billinge, J. Phys.– Condes. Matter 19 (2007). [22] G.L. Bykov, A.V. Gordeev, T.K. Yurik, B.G. Ershov, High Energy Chem. 42 (2008) 211–214. [23] K. Andersson, B. Allard, M. Bengtsson, B. Magnusson, Cem. Concr. Res. 19 (1989) 327–332. [24] R.R. Lloyd, J.L. Provis, J.S.J. van Deventer, Cem. Concr. Res. 40 (2010) 1386– 1392. [25] M. Kruk, M. Jaroniec, Chem. Mater. 13 (2001) 3169–3183. [26] K.S.W. Sing, D.H. Everett, R.A.W. Haul, L. Moscou, R.A. Pierotti, J. Rouquerol, T. Siemieniewska, Pure Appl. Chem. 57 (1985) 603–619. [27] L. Sarkisov, P.A. Monson, Langmuir 17 (2001) 7600–7604. [28] T. Tao, A.M. Glushenkov, Q. Chen, H. Hu, D. Zhou, H. Zhang, M. Boese, S. Liu, R. Amal, Y. Chen, CrystEngComm. 13 (2011) 1322–1327. [29] V. Kahlenberg, H. Bohm, Am. Mineral. 83 (1998) 631–637. [30] F. Piao, W.G. Oldham, E.E. Haller, J. Non-Cryst. Solids 276 (2000) 61–71. [31] A. Sandhu, S. Singh, O. Pandey, J. Phys. D Appl. Phys. 41 (2008) 165402.