Formation And Destabilization Of Organo-Aluminate AFm Phases

Transcription

1 Formation And Destabilization Of Organo-Aluminate AFm Phases 1 Giraudeau, C., d Espinose de Lacaillerie, J.-B 1 * 1 Ecole Supérieure de Physique et de Chimie Industrielles, ESPCI ParisTech (UMR CNRS 7615 Science et Ingénierie de la Matière Molle) Paris, France 2 Nonat A. 2 Université de Bourgogne (UMR CNRS 5209 ICB) Dijon, France 3 Flatt, R. 3 Sika Technology AG, Zürich, Switzerland Abstract The formation of organo-aluminate phases is a process speculated to consume superplasticizers, thereby reducing their efficiency as a dispersant in concrete. Over the past years various studies demonstrated that polycarboxylate polymers can be incorporated into organo-aluminate phases. We have shown that the interaction of the Ca-Al Layered Double Hydroxide, the so-called AFm phase, with anionic polycarboxylate ether based dispersant polymers (PCE for short) is not a simple adsorption but a more complex intercalation phenomenon leading to the transient sequestration of the PCE within the AFm crystallites. As a result, part of the PCE is immobilized forming a layered organo-mineral composite and does not play its role of dispersing agent. In a previous contribution, the prediction of scaling laws derived from models of conformation of comb copolymers in solution were tested against Small-angle x-ray diffraction (XRD) experimental results. A model of adsorbed polymers in a configuration of a flexible chain of hemispheric cores appeared compatible with the observed interlayer spacings in the range of several nanometers. Here, the chemical analysis of the intercalation solutions reveals that the intercalation mechanism probably proceeds though dissolution / precipitation over several hours. The release of the PCE in presence of sulfate ions also is not an immediate process and probably proceeds in the same way, though a dissolution / precipitation process. Originality The formation of intercalation composites between PCE's and hydrocalumite layers obtained by coprecipitation in a lime saturated solution containing the PCE was experimentally established. We were able to rationalize the composite structure using scaling laws derived from excluded volumes considerations for comb copolymers in solution. Chief contributions The PCE intercalated hydrocalumite composite was able to resist sulfate exchange for times in the range of hours. We thus propose that the PCE-AFm nanocomposite formation could be a practical key to understanding the dependence of superplasticizer effects on the cement composition, noticeably its calcium aluminate content. Keywords: Aluminates, ion-exchange, nanocomposites, polymers, Portland cement, superplasticizers 1 Corresponding author: jean-baptiste.despinose@espci.fr Tel , Fax

2 Introduction Ordinary Portland cement contains over 5% calcium aluminate in mass. It is the most reactive of the anhydrous phase and to maintain workability, its hydration must be controlled by gypsum addition. Careful studies of the early stages of hydration of C3A have established the conditions under which the hydroxy-afm can form as a transient phase leading eventually to sulfate-afm (monosulfoaluminate) or ettringite (trisulfoaluminate) (Minard et al. 2007; Pourchet et al. 2009). Nowadays, high-performance concretes can be formulated with a reduced water-to-cement ratio due to the admixture of polycarboxylate ether superplasticizers (PCE). The efficiency of these formulations however can be perturbed and the occurrence of rheological aberrations has been correlated to the aluminum content of the clinker (Flatt et al 2001). Two research groups, ours and the one of J. Plank, following an initial suggestion of V. Fernon (Fernon et al. 1997), have independently proposed that the intercalation process of PCE within the lamellar structure of AFm in presence of sulfate ions provides a key to the understanding of the seemingly erratic flow properties of some PCE cement admixtures (see for instance Giraudeau et al. 2009; Plank et al and references therein). After recalling the basic principles behind PCE intercalation, we will first briefly present the two intercalation models proposed by us (Giraudeau et al. 2009) and in Plank s paper (Plank et al. 2010). Then, we will study and discuss how the formation of PCE-AFm intercalate constitutes a short-term sequestration mechanism for the superplasticizer which can be unlocked by increased sulfate concentration as gypsum dissolution proceeds. Summary of PCE-AFm intercalation phenomenon The AFm family encompasses all the calcium aluminate hydrates based on the Ca/Al layered double hydroxide (LDH) structure. LDH s form when hydrolysable divalent and trivalent cations (respectively Ca(II) and Al(III) ions in the case of AFm) coprecipitate in brucite like layers. The presence of the trivalent cation results in an excess positive charge that is compensated by interlayer anions; the structural formula is thus [M II M x III (OH) 6 ] +.x/na n-.yh 2 O (Duan et al. 2006). In a cement paste, available anions are hydroxides, sulfates or carbonates thus forming the hydrocalumite, the monosulfoaluminate or the monocarbonate phases respectively. PCE s are poly(methacrylate-g-polyethylene oxide). At the high ph prevalent in a cement paste, all the carboxyl groups are deprotonated and they behave as polyanions. The basic hypothesis is that these polyanions can substitute the anions of the layered AFm structure to form organo mineral intercalates. This fact has been verified and reported several times in the literature on pure phases obtained either by coprecipitation or C3A hydration in the absence of carbonates. The signature of the formation of the organo mineral intercalate is the shift of the basal spacing observed by X-Ray Diffraction (XRD) to the high values needed to accommodate the bulky PCE anion. In some case, it has even been possible to visualize in direct space by Transmission Electron Microscope the high basal spacing of the intercalates. Actually, the basal spacings measured by small-angle XRD lead to unexpectedly high values ranging from 3 to about 12 nm depending on the structure of the PCE. This raises two fundamental questions. First, due to their high charge density, LDH layers, contrarily to clays, are known not to spontaneously delaminate in water, so, how can their layers be propped apart to accommodate the PCE anions? Second, how can they swell in the range of 5 nm and still maintain coherence? In a previous paper (Giraudeau et al. 2009), we proposed to rationalize the interlayer spacing by the steric effect of the polymer due to a blob conformation model derived from the grafted polymer conformations in solution (Gay and Raphaël 2001). Our model is represented in a schematic form in Figure 1.

3 Figure 1: Author s model of polymer conformation in the interlayer space of hydrocalumite as a bilayer of chains of hemispheric cores. This model predicts an interlayer of twice the hemispheric radii as indeed observed by small angle XRD (see (Giraudeau et al. 2009)). Plank et al proposed instead that the polymer maintains a comb configuration and that the high basal spacings result from the formation of staged intermediates with sulfate ions (Figure 2) (Plank et al. 2010). Indeed, staged intermediate have been shown to occur in other LDH compounds (Fogg et al. 1998). On the one hand, this model is satisfactory in the sense that it allows for the observation of high basal spacings while keeping the distances between the adjacent LDH layers within the range of what is commonly observed, namely, a few nanometers at most. Furthermore, it provides a credible picture for polymer exchange without delamination since it is the flexibility of the LDH layer that permits diffusion of the polymer within the inter-lamellar space. Figure 2: Plank s model of polymer conformation in the interlayer space of hydrocalumite as sulphate / comb polymer second stage intercalation compound (reproduced from Figure 7 in (Plank et al. 2010)). The arrow shows the large basal distance (up to 7 nm) which, in the case of a second-stage intermediate, includes two interlamellar spaces.

4 On the other hand, the high basal spacings can be observed in the absence of sulfate ions as evidenced in our previous study. In addition, in the same study, the sizes of the hydrocalumite crystallites under consideration were in the range of 5 to 9 nm, corresponding to a surface of 20 to 80 nm 2 for each side of each LDH platelets, while, for the PCE s used, the chains extended over a surface of 50 to 120 nm 2. Therefore, even in a bilayer intercalation model, there was just about the right amount of LDH surface to accommodate the polymer. Retrospectively, it thus appears that this fact is not compatible with a staged structure which requires an excess of LDH surface respective to the surface covered by the polymer. In conclusion, the high basal spacings we observed in the absence of sulfate did not correspond to the staged compound evidenced in the presence of sulfate by Plank s group. In our opinion, the issue thus remains partially unresolved. To further address the issue, it is of interest to address the formation mechanisms of the intercalated PCE / AFm structure. We thus decided to carry a kinetic study by following the ions concentrations with time after PCE and sulfate addition to a pre-formed OH-AFm phase (hydrocalumite). Experimental The hydrocalumite suspension was synthesized following the same procedure as in our previous study (Giraudeau et al. 2009). The end result is an aqueous suspension of 4.0 mm [Ca 2 Al (OH) 6 ] + nanoplatelets. It is kept under Argon to avoid carbonatation and maintained at 5 C to prevent the formation of katoite. 2g / L of polymers were then added and the Al and Ca concentrations in the solution monitored at various time intervals by ICP-AES after filtering out the solid through a 0.3 µm filter. The concentration of 2 g /L was chosen to mimic the polymer concentration in a cement paste after addition of 0.1% weight ratio of polymer to cement for an e/c ratio of 0.5. The evolution of the polymer concentrations in solution was followed by Total Organic Carbon analysis. After 4 days, Na 2 SO 4 was added in the proportion needed to obtain the stoichiometry of monosulfoaluminate, namely SO 4 2- /Al 3+ = 0.5. Sulfate concentrations were measured in solution by ICP-AES. The general formula of the PCES is given in Figure 3. Their characteristics are given in Table 1. Figure 3: General formula of the polycarboxylate ethers used in this study. The values of N, x and p are given in Table 1.

5 Table 1: Characteristics of the PCE s used and time constant obtained from the kinetic study of intercalation. Backbone chain length (N) Side chain length (p) Charge per unit (x) A h C h D h E h Intercalation time constant τ Results and discussion To gain some insight, we performed a kinetic study of the PCE-AFm composite formation by analyzing the concentrations in solution with time after addition of the polymer. Not surprisingly, the intercalated polymer content increased exponentially to an equilibrium value according to the equation (Figure 4): C intercalated = C equilibrium [1 - exp(-t / τ)] (1) All the PCE s of the present study followed the same trend with time constants τ in the range of hours (Table 1). It is remarkable that the intercalation took several hours to reach equilibrium, indicating a significant energy barrier to intercalation. As discussed previously, depending on the intercalation model, this barrier can be circumscribed by the flexion of the layers necessary for staging, or by the dissolution of the hydrocalumite layers before precipitation of the intercalated composite. Monitoring the concentration in solution of Al and Ca, the constitutive ions of the hydrocalumite layer provided an interesting clue. Initially, the Al and Ca concentrations are not zero as they reflect the solubility of the hydrocalumite. The addition of the polymer triggers a transient increase of their concentrations as seen in the case of PCE E in figure 5. The maximum concentration of inorganic ions is obtained after 2 h, which is precisely when the adsorption of the PCE E becomes significant. This shows that the composite formation mechanism contains several successive steps, one of them occurring in solution. When adding sulfate ions to the PCE-AFm suspension, the intercalated PCE is released into solution as previously reported (Giraudeau et al. 2009). For brevity, the data are not reproduced here but, again, the release of the PCE is not immediate and it takes 2 to 5 hours depending on the PCE structure for the PCE to be released in the solution.

6 0,5 1,2 1,0 Adsorbed PCE (g / L) 0,3 0,1 A : τ = 3.2 h Adsorbed PCE (g / L) 0,8 0,6 C : τ = 2.7 h 1,0 0,8 0,8 Adsorbed PCE (g / L) 0,6 D : τ = 2.9 h Adsorber PCE (g / L) 0,6 E : τ = 5.6 h Figure 4: Evolution with time after addition of PCE to a hydrocalumite suspension of the amount of intercalated PCE s. The values of τ are obtained by fitting with the equation given in the text (solid line). [Al] 3+ (x10) and [Ca] 2+ in solution (mmol/l) Ca Al (x10) Time (h) Figure 5: Evolution with time after addition of PCE E to a hydrocalumite suspension of the amount of Ca and Al in solution. The Al concentration is multiplied by 10 to improve the readability of the figure.

7 Conclusion Previously reported models of PCE / AFm intercalation (staged intermediate or blob intercalation) require circumscribing the propping apart of the hydrocalumite layers. However, LDH carry a high surface charge density (about 1.75 positive charges per nm 2 ) and do not spontaneously delaminate in water. Consequently, the kinetics of intercalation is expected to be slow. In this study, we report that indeed, intercalation can take several hours to reach equilibrium. Furthermore, we show that during this time-lag, the concentration of Al and Ca in solution are not constant, suggesting that the high energy barrier of delamination is circumscribed through a dissolution/precipitation mechanism. The same time lag, and thus the same mechanism, is expected for the PCE release in presence of sulfates. Acknowledgments This study was funded by the NANOCEM consortium and a CIFRE grant from the French government. References Duan, X. and D. G. Evans, Eds. (2006). Layered Double Hydroxides. Structure and Bonding. Berlin, Springer. Fernon, V., A. Vichot, et al. (1997). Interaction between Portland cement hydrates and polynaphtalene sulfonates. Fifth CANMET/ACI Conference on Superplasticizers in Concrete. Flatt, R. J. and Y. F. Houst (2001). "A simplified view on chemical effects perturbing the action of superplasticizers." Cement and Concrete Research 31(8): Fogg, A. M., J. S. Dunn, et al. (1998). "Formation of Second-Stage Intermediates in Anion-Exchange Intercalation Reactions of the Layered Double Hydroxide [LiAl2(OH)6]Cl.H2O As Observed by Time-Resolved, in Situ X-ray Diffraction." Chemistry of Materials 10(1): Gay, C. and E. Raphaël (2001). "Comb-like polymers inside nanoscale pores." Advances in Colloid and Interface Science 94(1-3): Giraudeau, C., J. B. d'espinose de Lacaillerie, et al. (2009). "Surface and Intercalation Chemistry of Polycarboxylate Copolymers in Cementitious Systems." Journal of the American Ceramic Society 92(11): Minard, H., S. Garrault, et al. (2007). "Mechanisms and parameters controlling the tricalcium aluminate reactivity in the presence of gypsum." Cement and Concrete Research 37(10): Plank, J., D. Zhimin, et al. (2010). "Fundamental mechanisms for polycarboxylate intercalation into C3A hydrate phases and the role of sulfate present in cement." Cement and Concrete Research 40(1): Pourchet, S., L. Regnaud, et al. (2009). "Early C(3)A hydration in the presence of different kinds of calcium sulfate." Cement and Concrete Research 39(11):