In situ platelets formation into aqueous polymer colloids: The topochemical transformation from single to double layered hydroxide (LSH–LDH) uncovered

Journal of Colloid and Interface Science 462 (2016) 260–271

Contents lists available at ScienceDirect

Journal of Colloid and Interface Science journal homepage: www.elsevier.com/locate/jcis

In situ platelets formation into aqueous polymer colloids: The topochemical transformation from single to double layered hydroxide (LSH–LDH) uncovered Thomas Stimpfling a, Arthur Langry a, Horst Hintze-Bruening b, Fabrice Leroux a,⇑ a b

Université Clermont Auvergne, Université Blaise Pascal, Institut de Chimie de Clermont-Ferrand – ICCF, UMR-CNRS n°6296, BP 80026, F-63171 Aubière, France BASF Coatings GmbH, Glasuritstrasse 1, 48165 Muenster, Germany

g r a p h i c a l a b s t r a c t

a r t i c l e

i n f o

Article history: Received 27 August 2015 Revised 2 October 2015 Accepted 5 October 2015 Available online 9 October 2015 Keywords: Layered structure Topochemical transformation Polymer nanocomposite Platelets intercalation versus exfoliation Solid state kinetics

a b s t r a c t Layered Single Hydroxide (LSH) of chemical composition Zn5(OH)8(acetate)2
nH2O is synthesized under in situ condition in an aqueous dispersion of an amphiphilic, carboxylate bearing polyester via a modified polyol route. The one-pot LSH generation yields agglomerates of well intercalated platelets, 9–10 nm separated from each other. However the corresponding Layered Double Hydroxide (LDH) of formal composition Zn2Al(OH)6 (acetate)
nH2O is found to proceed via the formation of crystallized, similarly spaced LSH sheets in the neighborhood of amorphous Al rich domains as evidenced by X-ray diffraction and transmission electron micrographs. The initial phase segregation effaces over time while LSH platelets convert into the LDH phase. Fingerprinted by the change of in-plane cation accommodation, the associated topochemical reaction of the edge-sharing octahedral LSH platelets involves the transformation of metal lacunae, adjacently covered by one tetrahedral coordinated cation on each side to balance the negative surcharge, into fully occupied and monolayered platelets of edge-sharing octahedral LDH, the former voids being occupied by trivalent cations. This replenishing process of empty sites, coupled with the dissolution of tetrahedral sites is likely to be observed for the first time due to the presence of well separated, polymer intercalated platelets. TEM pictures vision crystal growth arising from the zone of the LSH edge-slab and by using solid state kinetics formalism the associated high activation energy of the first-order reaction agrees well with a plausible dissolution re-precipitation mechanism. The conversion of LSH into LDH platelets may be extended to others cations as Co2+, Cu2+, as well as the aluminum source (AlCl3) and the water-soluble polymer (NVP), thus indicating it is a new prevalent facet of LDH. Ó 2015 Elsevier Inc. All rights reserved.

1. Introduction Layered materials are ubiquitous in nature, resulting from pressure and mineral dissolution–precipitation reactions. Rarely homo⇑ Corresponding author. E-mail address: [email protected] (F. Leroux). http://dx.doi.org/10.1016/j.jcis.2015.10.010 0021-9797/Ó 2015 Elsevier Inc. All rights reserved.

geneous, they usually display stacking default or even interstratification. Such structural modifications implying polysomatic order – disorder also called mixed layering are well known in clay minerals, e.g. trioctahedral micas, chlorites, illites. For the latter this is illustrated by repetitive tetrahedron–octahedron– tetrahedron (TOT) layers [1]. In laboratory a lot of attention is devoted to design inorganic layered material because of the

T. Stimpfling et al. / Journal of Colloid and Interface Science 462 (2016) 260–271

associated structural anisotropy that makes them attractive as barrier, container and high modulus-bearing fillers for polymers and coating systems [2]. Indeed, the possibility for some layered systems to intercalate guest molecules renders them attractive as reservoirs, capable to act upon chemical stimulus and as a delivery system [3]. We focused our attention on Layered Double Hydroxide materials for which tremendous effort has recently been paid to fabricate nano-sheets as building blocks [4–8] and subsequently to use them in some kind of nano-engineering approach [9,10], making possible their combination with other materials of high aspect ratio such as graphene [11], layered titanium oxide yielding mesoporous nanohybrid layer-by-layer ordered [12] or carbon nanotubes [13]. To some extent such a combination is also possible with three-dimensional materials as reported for metal–organic framework (MOF) membranes [14]. A possible strategy to enhance interfacial interaction between 2D-sheets and organic matrix materials like polymers is to generate one of both constituents in the presence of the other as a medium of choice. Indeed, rather than exfoliation, we were interested in generating LDH nano-sheets within a continuous polymer medium following the idea to hybridize directly the resulting nanolayered architecture. However such synthetic in situ approach may yield metastable phases. There is lot of interest to understand possible transformations of layered inorganic systems. This prompts us to study the generation of Layered Single Hydroxide (LSH) and Layered Double Hydroxide (LDH) and as a critical issue in term of phase stability to address whether one system might be involved in the synthesis of the other. The layered structures of both LSH and LDH are displayed in Scheme 1. Hydroxy Double Salt (HDS) also called Layered Double Salt (LDS) or Layered Mixed Basic Salt (LMBA) were first reported for layered hydroxide metal acetates (also referred as layered double acetates) adopting the general formula M(OH)2x(CH3COO)x
nH2O with M being a divalent cation [15]. Zinc basic salts Zn5(OH)8 (NO3)2
2H2O were discovered a long time ago [16], and consist of zinc based brucite-type hydroxide layers comprising vacancies where a zinc ion occupies a tetrahedral site both above and below each vacancy. LDH materials, also called hydrotalcite-type compounds can be II described as follows: [MII1xMIIIx(OH)2]+x[Ax/m m
nH2O] where M ,

261

MIII and A are the divalent, trivalent and interleaved anion, respectively [17]. Both LSH and LDH materials present: – Quasi-planar array of edge-sharing octahedral, complete in the case of LDH but incomplete for LSH, where metal vacancies are charge balanced by two tetrahedral M(II) sites, one adjacently attached on each side of an octahedral void, and as much as one quarter of the octahedral sites of the metal hydroxide layers can be vacant at maximum. – Anions interleaved between the layers to compensate an excessive positive charge stemming from a trivalent cation in the case of LDH or to coordinate a tetrahedral divalent cation in the case of LSH, respectively. LDH formation attracted attention and four different mechanisms have been proposed: (i) coprecipitation occurs for hydrotalcite composition through first the formation of amorphous colloidal aluminum hydroxide then transformed into the lamellar oxide-hydroxide aluminum boehmite c-AlOOH in which surrounding Mg2+ cations are incorporated leading to a charge imbalance of the sheets [18]. Such precipitation steps involving classically the formation of hydroxide/hydrous oxide intermediates observed in the case of the cations Al3+ into hydrotalcite-type framework [19,20] may evidently differ for other metal cations, as for Cr3+ where it proceeds to form a Cu2Cr-type LDH material through a short-range cation order [20] confirmed by the formation of Cr oligomers and hexa-aquozinc whose condensation yields cationic order [21] – (ii) topotactic reaction through anion exchange reaction from a template consisting in a pristine LDH material possessing easily exchangeable anions [22] – (iii) dissolution – crystallization for reconstruction [23] – (iv) dissociation–deposition– diffusion for hydrothermal synthesis [24]. For LSH formation mechanisms have not been reported yet but the synthetic approaches do not differ from those for LDH with direct precipitation, hydrothermal treatment [25] and through anion exchange [26,27]. Lately micro-wave assisted synthesis was added [28]. Rather ubiquitous LDH-type platelets have also been found to be the product of another structural evolution: dissolved divalent cations reacting with an aluminum source like aluminum hydroxide or c-alumina as a solid substrate in an aqueous solution. Pioneered using Zn2+ [29], Ni2+ and Co2+ [30,31], it was recently

Scheme 1. Ideal structure of (a) LSH and (b) LDH based on Zn5(OH)8(CH3COO)
nH2O and Zn2Al(OH)6(CH3COO)
nH2O, respectively.

262

T. Stimpfling et al. / Journal of Colloid and Interface Science 462 (2016) 260–271

shown that the reaction involves local dissolution of the substrate [32], followed by precipitation at the solid liquid interface. Thus partial dissolution of c-alumina forms Al(OH)3 on which M2+ ions are adsorbed, evidenced by a polyhedral change from AlO6 to AlO4 in the zinc aluminate case [33]. The amphoteric nature of Al3+ cations was underlined to explain why LDH-type layers ultimately form from other phases and substrates [34]. While intercalation in a solid layered structure is a topotactic reaction leaving the platelets intact, the intra-sheet cation accomodation can be defined more as a topochemical reaction where the chemical conversion of a starting solid yields a reactive product, the progress of the conversion being connected to a distinct site in the crystal. Pioneered by Kohlschütter [35] and Feitknecht [36] renewed by Boldyrev [37], the term topochemistry (from Greek sopor meaning site) is now largely employed in materials science. It defines conversion mechanisms where the transformation is involving reactive sites associated to diffusion at the interface although leaving intact the whole appearance. The conversion is usually triggered by soft redox reaction as illustrated in the case of 2D-ferromagnetic oxide BaFe2(PO4)2 [38], the transformation of Mn(OH)2 into layered manganese oxide [39], or the formation of alkali-metal hydroxide layers within double- and triple-layer perovskites [40]. Concerning LDH-type framework, a direct correlation between nanostructure and particle morphology is observed through surfactant exchange reaction leaving constant the aspect ratio [41]. For the intra-layer arrangement, a homothetic conversion of the edge-sharing octahedral array is reported to occur from brucite b-Co(OH)2 [42–44]. Such structural/chemical evolution leaving intact bidimensionality can be also defined as a topotactic transformation and it proceeds from an outside-source through cation migration rather than through the material itself. Indeed the resulting phase should inherit the general aspect from the pristine phase through this kind of morphosis. Thus gibbsite c-Al(OH)3 with 1/3 of empty octahedral sites is well known to imbibe lithium salts thus transforming into LDH-type layers of the cation composition LiAl2 which have been shown to crystallize in an eclipsed stacking along the c axis [45,46]. A mechanism called ‘‘diadochy” has been reported to explain the selective cation intra-sheet substitution of Mg2+ by transition metal cations, but it concerns very few cases and low substitution rate [47]. In both cases, the metal sites maintain their local octahedral coordination symmetry after the metal uptake or exchange respectively. Here a coprecipitation using the polyol route was adopted. It consists in the slow hydrolysis of metal acetate salt dissolved in an alcoholic medium by using aqueous ammonia and which was found to efficiently form both LSH [48] and LDH [49] platelets. Formed acetate-tactoids of cation composition Ni:Al were reported to undergo facile aqueous exfoliation through an ordered interstratified phase consisting in hydrated and dehydrated layers stacked alternatively and resulting in a staged S-2 phase [50]. Additionally it was demonstrated that acetate anion presents an extremely low anion selectivity toward LDH phase [51], this aspect could be of interest in producing exfoliated layers as reported in the case of Ni:Al composition [50]. Environmental benign, inexpensive and easily removable, acetate anions render this route an attractive approach in the formulation of 2D platelets based polymer composites. In the presence of water dispersible amphiphilic polyester, stabilized via carboxylate groups, LSH-type layers are formed, however in presence of an aluminum source LSH platelets progressively fade over time in favor of LDH. This LSH into LDH conversion within an Al-eutrophic medium is found to be highly temperature dependent regarding the rate. Therefore, we aimed to describe this conversion phenomenon by using solid-state kinetic models as well as the Arrhenius equation in order to

determine the mechanism of the process and its activation energy, respectively. Structural analyses were performed under different conditions of aging times and temperatures. Finally the systems studied were extended to other components: from polyester to poly(vinyl)pyrrolidone (PVP) for the water-soluble polymer, from Al3+(hydroxyl)acetate to Al3+ chloride for the Al source and finally from Zn2+ to Co2+ or Cu2+ for the divalent metal. 2. Experimental section 2.1. Chemicals Zinc acetate dehydrate (Zn(CH3COO)2
2H2O, Sigma–Aldrich), Aluminum acetate basic (Al(CH3COO)2(OH), Sigma–Aldrich), Cobalt acetate (Co(CH3COO)2
4H2O, Sigma–Aldrich), Cupper acetate (Cu(CH3COO)2
H2O), ammonium hydroxide 25% (Sigma– Aldrich) and polyester (as hereafter called PESh 60 wt% solution in a 1:1 parts by weight mixture of 2-butoxyethanol and water, as described previously [52]), poly(-N-vinylpyrrolidone ((C6H9NO)n, noted as PVP, M.W. = 8.000 g mol1, Acros) were used as received. 2.2. Samples preparation Single and double layered hydroxides were prepared according to a modified polyol route. M2+ and M2+, M3+ salts were dissolved into 125 mL of 2-butoxyethanol at 80 °C before hydrolysis with 50 mL deionized water at pH adjusted between 7 and 8 by addition of 10 mL of ammonia. The mixture was then aged at 80 °C overnight following by three successive washing steps and centrifugation at 4000 rpm. The resulting paste was dried in an oven at 40 °C for 24 h (yield for LSH and LDH was of 0.67 and 0.68, respectively). From 6.7 g of the as-received polyester (PES) solution (60 wt%) diluted in 50 mL of distilled water, 100 mL of 0.18 M solution of the divalent salt (M2+ = Zn2+, Co2+, Cu2+) or the mixed M2+, Al3+ salts with a M2+/Al3+ ratio of 2 respectively were added dropwise at room temperature. Such mass balance between PES and inorganic platelets was chosen for a theoretical filler content of 10 wt% in the case of an in situ synthesis of complete yield. The addition was performed at a constant flow of 0.33 mL/min under nitrogen atmosphere. The pH value was maintained at a constant value of 9 ± 0.1 via the addition of 0.65 M of NH4OH solution. One should note that no washing procedure was applied. At this step referred as starting time t0 and in absence of washing and drying, the slurries were analyzed by XRD and SAXS (see below) or kept under air in sealed flask at different temperatures for further analyses upon time. Synthesis using PVP was accomplished as previously reported using a 50 mL diluted aqueous solution where 10 g of PVP was dispersed. In all cases, the slurries were rather inhomogeneous and became more and more viscous over time. Mixtures were found to be more homogeneous using PVP, for which gel-like suspensions eventually were obtained. For extended storage, after successively having taken samples, a skin formed at the surface of the mixtures. Therefore samples were always carefully taken out of the bulk, liquid system. We surmise further that few mg samples taken from the ageing batch can be neglected as a major change of the overall composition as well as a significant disturbance of the transformation kinetics. Experimentally at a given time, a certain amount of paste (10–20 mg) was isolated using a spatula and dried at room temperature for analysis. It was checked by XRD analysis that the structural evolution was quenched by the drying process, i.e. once dried the specimens did not evolve further in time. Identical methodologies were applied for all the systems using all the M2+ as well as for the mixture using PVP. A step of centrifugation was possible for low viscous slurries to speed up the drying process.

T. Stimpfling et al. / Journal of Colloid and Interface Science 462 (2016) 260–271

2.3. Techniques of characterization Powder X-ray diffraction (XRD) profiles were obtained with Siemens Model D500 X-ray diffractometer with diffracted beam monochromator and Cu Ka source. Patterns were recorded over the 2h range of 2–70°, in steps of 0.08° with a count time of 4 s. Ultrathin sections (70 nm) of dried and time-stabilized specimens of polyester LSH–LDH nanocomposite were cut at 80 °C using a Leica UC6 ultra-microtome equipped with a Leica EM FCS cryo chamber and a diamond knife. Cryogenic transmission electron microscopy (cryo-TEM) was performed on a FEI Tecnai G2 electron microscope using 200 kV as accelerating voltage. Small Angle X-ray Scattering (SAXS) was performed at the synchrotron facility SOLEIL using the beamline SWING. Long distance correlation was displayed between 0.00327 and 0.327 Å1 in the q-domain, i.e. distances ranging from 19 to 190 nm, while the shorter distance correlation was displayed from 0.02168 to 1.6269 Å1, i.e. 0.39–29 nm. Both were recorded using a distance between detector of PCCD170170 (AVIEX) type and sample of 2629.006(1) and 505.021(8) mm, respectively. Time recording took 5 and 50 ms for the long and shorter distance correlation, respectively, and 10 spectra were accumulated to get a proper signal to noise ratio. Temperature was applied between 8 °C and 65 °C on freshly prepared slurries to reproduce the thermal dependence in the early ages. For aged and dried specimens, the spectra were recorded at room temperature. 3. Results and discussion 3.1. Acetate-LSH (LZH) and its in-situ formation into water-soluble polymer LSH/acetate material prepared by coprecipitation via the polyol route described by Poul et al. [48], using 2-butoxyethanol as solvent, in the presence of the water-soluble polyester exhibits a simple X-ray pattern (Fig. 1a). A representative XRD pattern consists of narrow diffraction lines close to 2h = 33° and 59° and a large hump centered at 2h = 20°. The latter is arising from the presence of polyester, while the former are indicative of in-plane ordering of a LSH-type material in agreement with the literature [48] and thus suggest the formation of inorganic LSH layers. As the number of diffraction lines is rather limited, LSH material is prepared as previously described without the water-soluble polymer to confirm

263

the presence of LSH platelets. Displayed in Fig. 1b, the expected X-ray pattern of LSH/acetate, Zn5(OH)8(acetate)2
nH2O material is  rhombohedral symmetry, shown. The structure is described in R3m showing the presence of harmonic peaks (0 0 l) at lower 2h values and the (1 1 0) diffraction line close to 59°. The lattice parameter a obeys the law of Bragg and corresponds to the average in-plane cation–cation distance which causes the peak at 59.28°, attributed to the (1 1 0) diffraction. A cell parameter a of 0.312 nm agrees with Zn5(OH)8(acetate)
2H2O structural data. The presence of (0 0 l) harmonics correspond to a basal spacing of 1.62 nm, slightly higher than the 1.47 nm reported for a monophasic LSH (Zn) acetate phase [48], but having in mind a shift around this value may be possible due to the orientation of acetate groups [53]. Interestingly, such interlayer information is absent when the synthesis is performed in the presence of the polyester, thus suggesting that the stacking of the platelets is not readily occurring. To the best of our knowledge this would be a novelty for LSH type materials that are known to expand their gallery height not exceeding a few tenths of nm for instance from 1.472 to 2.034 nm by ethylene glycol [54]. With regard to polymer processing only poorly dispersed microcomposites of poly(methyl)metacrylate intercalated LSH platelets have been reported [55,56]. When dried and further cryo-prepared for TEM observation and this for the shorter storage time (closest on the time scale regarding the samples condition to record Fig. 1a), platelets of an apparent thickness of 1–2 nm are observed and organized as homogeneously dispersed bundles of layers (Fig. 2). These are slightly corrugated and laterally extending to approximately 200 nm, their average distance being roughly 9–10 nm. This corresponds to 2h values as low as 0.88–0.98° which is beyond the XRD range. The visible hump in Fig. 1a at around 2h = 6.15° may correspond to LSH acetate platelets stacked in small number [48], or a harmonic peak of higher order (around 7th order) indicating that apparently invisible for XRD pronounced stacking occurs, thus confirming the TEM pictures showing extended domains of stacked fringes (Fig. 2). To further scrutinize the lamellar periodicity present at the early times of LSH synthesis within the polymer, small-angle X-ray scattering is performed on fresh samples prepared in the same way as before. The temperature of the samples was let to vary in the range of 8–65 °C. The basal spacing is found to increase from 18 to 31 nm (Fig. 3). The slope of the scattering intensity as a function of the scattering vector is close to 2 (2.1), consistent with the presence of platelets. A large platelets expansion up to 20-fold the initial basal spacing is reminiscent of gigantic swelling observed with protonated titanate in aqueous solution [57]. Indeed the discrepancy in spacing values between USAXS results obtained on fresh samples that were not allowed to dry like those used for cryo-preparation might reflect the swelling of a periodically layered system in a low viscous solvent/polymer system and its collapse in the course of solvent evaporation. The increasing spacing of the swollen system with increasing temperature might be associated with a gain of entropy of the confined polymer chains and solvent molecules.

3.2. LDH composition into water-soluble polymer

Fig. 1. X-ray pattern for LSH materials obtained from polyol route (a) in presence and (b) in absence of the water-soluble polyester. For (b0 ) y-intensity of (b) was multiplied by 6 for a better reading. Reflection lines (hkl) characteristic of intraplane arrangement are materialized.

When applying the in situ mode to prepare LDH platelets, the obtained XRD pattern largely presents similarities with the previous in situ synthesis (Fig. 4a). Indeed the shapes as well as the position of the diffraction lines are similar to those of the LSH in situ preparation, suggesting that the formed platelets are LSH rather than LDH. To exclude a possible coexisting Zn2+ and Al3+ occupancy of the octahedral sites in the layered hydroxide, LDH/acetate is

264

T. Stimpfling et al. / Journal of Colloid and Interface Science 462 (2016) 260–271

Fig. 2. Typical TEM images of LSH platelets at different scales indicated on the bars showing the layered structure arranged as bundles and forming aggregates.

Fig. 3. SAXS spectra for LSH platelets. The temperature is indicated.

prepared as before via the polyol route using 2-butoxyethanol as solvent for comparison [49]. X-ray data (Fig. 5) are characteristic  rhombohedral for a LDH structure, that is usually described in R3m symmetry, and the (1 1 0) diffraction line at 61.05° corresponds to a cell parameter of a = 0.304 nm; a small though significant shift in the angular values 2h  1° (KCu) according to one hundredth of a nm in term of distance (cf. a = 0.312 nm for LSH). Variation of the diffraction peak (1 1 0) position is known to result from

Fig. 4. X-ray patterns of LDH/samples prepared in situ in presence of the watersoluble polyester after an aging time of (a) t = 0 h, and after (b) 170 days. The dashed lines materialize the main intralayer diffraction peaks different between LSH and LDH sheets as identified in Fig. 1.

stoichiometric change even of small amplitude as for closely related cations [58]. This is accompanied with a broadening of LDH diffraction peaks. Sharper diffraction lines for LSH have to be explained on the basis of similar sizes regarding occupied (Zn(OH)6) versus empty octahedral sites (h(OH)6) compared with the in-plane structure of LDH which has to accommodate two

T. Stimpfling et al. / Journal of Colloid and Interface Science 462 (2016) 260–271

Fig. 5. Typical X-ray pattern of acetate phases for (a) Zn-LSH and (b) Zn2Al-LDH prepared by the polyol route (in absence of polymer). The dashed lines materialize the main reflection peaks informing on the intra-layer arrangement different between LSH and LDH sheets as identified in Fig. 1.

cations of different ionic radius. This results in biased Zn2+(OH)6 and Al3+(OH)6 octahedra as a consequence of their edge-sharing connectivity. Thus the average distance between the cations is a function of their ratio and the relative content of alumina cations may be extrapolated from the dependence of the cell parameter a through the relationship

da=dx ¼

p

3þ

2frðZn2þ Þ  rðAl Þg;

where r(Zn2+) and r(Al3+) are the ionic radius of the respective 3+ cations, and x the degree of substitution in Zn2+ 1xAl x(OH)2. The 2+ 3+ ionic radius for Zn of 0.074 nm and for Al of 0.054 nm, results in a dependence da/dx of 0.0325 nm per x. A value of 0.304 nm for the cell parameter a associated to the cation relative composition 3+ Zn2+ 0.66Al0.33 can be considered as a reference position for widespread stoichiometry of LDH materials with M2+/M3+ = 2 [59]. Thus a cell parameter of 0.312 nm would correspond to x = 0.084, resulting in a ratio Zn/Al of around 10.9. This value is much higher than the usual range of 2–4. It may be deduced that the experimental value is consistent with the presence of LSH platelets rather than a highly improbable Zn2+Al3+ LDH composition. The position and shape of other diffraction peaks reinforce the assumption of LSH formation.

265

broadening of the diffraction lines for the intra-layer structure and a shift toward the position typical for LDH platelets (cf. Fig. 5) were progressively found, as evidenced by Fig. 4b after 170 days. As far as intra layer substitution is concerned, only few examples are known such as the formation of LiAl2(OH)6(CO2 3 )
2H2O LDH type framework starting from gibbsite, where Li+ cations diffuse into the octahedral empty sites of the c-Al(OH)3 inorganic template [46]. However LSH platelets present tetrahedral coordinated zinc cations pointing out of the slab above and below unfilled octahedral sites. The LSH to LDH conversion thus not only involves the filling of empty octahedral sites but also the disintegration of the Zn2+ tetrahedral sites, as well as local disposal of the excessive Zn2+ cations. To gain more insights into the structural conversion, the ongoing transformation was progressively followed from early time to various shelf-life times stored at room temperature by recording XRD patterns which permits to visualize the conversion between both platelets systems (Fig. 6). As expected, the position of (1 1 0) diffraction line clearly shifts from 2h  59° to 61° (see enlarged 2h domain at high 2h values), fingerprinting the LSH to LDH conversion. On dried and cryo-prepared samples, the mechanism affecting the platelets dispersion in the polyester comprising reactive system was scrutinized by TEM at different times (Fig. 7). In fresh samples, Zn- and Al-rich domains are separated, the former comprising a layered structure while the latter consist of ill-defined aggregates as evidenced under higher magnification. In more detail, the layered structure presents similarities with the LSH-based mixture (Fig. 2) regarding stacked layers having a basal spacing of 9–10 nm. However the number of layers per bundle apparently is less compared to that of the LSH system. This direct observation agrees with the XRD pattern of polymer intercalated LSH platelets, with the h k 0 and 0 k l reflections displayed while the 0 0 l reflections are out of the measuring range, besides the amorphous, XRD ‘‘silent” Al source. Based on the average of several observations and not on the very same spotted sections, the initial relatively dense Al-rich domains progressively erode upon aging and apparently may even fragment into smaller pieces (see 0 h vs. 120 d in Fig. 7a). At their surface a needle like morphology progressively appears over time which could be attributed to the cross sectional view of platelets that have been grown as a shell on the

3.3. Evolution of the in-situ LDH dispersion A shift in the diffraction line (1 1 0) from 59.28° to 61.05° is observed. This behavior is evidently not expected from a topotactic reaction but has to be interpreted by a structural in-plane rearrangement of the LSH platelets. The corresponding final value is close to what is observed for LDH phase. The change in hk-dependent indices suggests a modification of the intra lamellar domain resulting from different bonding distances, i.e. cation 3+ accommodation, with a cation composition Zn2+ 0.66Al0.33 (vide supra). As no special care of storage was applied, in particular with regard to the atmosphere, the well-known ready uptake of carbon dioxide and intercalation of carbonate ions into the inorganic structure would be anticipated which would result in a welldefined stacking exhibiting a first-order harmonic diffraction peak at 2h = 10°. However carbonation reaction was found to be absent over prolonged storage time of 170 days and even after addition of Na2CO3. The latter addition provokes the degradation of the platelets into zincite (ZnO) (not shown). However when stored a

Fig. 6. Typical XRD patterns stacking on time at 50 °C from below to top t = 0, 7, 10, 15, 21, 25, 30, 35, 75, 80, 85 days. The dashed lines materialize the main diffraction peaks different between LSH and LDH sheets as identified in Fig. 1.

266

T. Stimpfling et al. / Journal of Colloid and Interface Science 462 (2016) 260–271

Fig. 7. Typical TEM images showing the structural transformation at different aging times at (a) Al3+ cation reservoir, (b) layered structures and (c) interfacial transitional zone. Major element composition is indicated.

T. Stimpfling et al. / Journal of Colloid and Interface Science 462 (2016) 260–271

267

Fig. 7 (continued)

domains surface (Fig. 7) while higher magnification reveals a fringy interphase (see evolution between 56 and 120 days in Fig. 7c). Over the whole period but more pronounced for longer aging times, layered structures appear in the zinc rich domains as needles which we interpret again as cross sectioned platelets (Fig. 7b). 3.4. Mechanistic aspect of the structural evolution To understand the conversion between LSH and LDH platelets, classical models interpreting the solid state kinetics were considered based on nucleation and nuclei growth models (i.e. ingestion or coalescence), dimension-dependent diffusive model, dissolution/ precipitation or simple order-based models when the rate depends on the concentration or remaining fractions of reactants [60]. Models of solid-state kinetics are known to be well adapted for the understanding of mechanisms involving layered systems like LDH phases [60,61]. From the angular domain zoomed from 55° to 65° in 2h-range, a shift of the diffraction line (1 1 0) is observed which reflects the LSH to LDH phase conversion. A representative deconvolution is displayed in Fig. 8a, and by using two Gaussian curves centered on each diffraction line (1 1 0) of both platelets systems, the ratio between them, i.e. LSH and LDH may be addressed. In details, the fraction of LSH to be converted into LDH (aLDH) at a certain time was estimated from the peak ratio between the areas of the Gaussian curve corresponding to the diffraction line (1 1 0) of LDH reported to the total area. The form of the function aLDH = f(t) is significant with regard to growth mechanism and can generally be described using the Johnson– Mhel–Avrami–Kolmogorov (JMAK) equation [61]: n

lnð1  aLDH Þ ¼ ðktÞ ; in which k is an effective rate constant and n the Avrami exponent related to the kinetic model.

Fig. 8b represents conversion values obtained at different temperatures over time, fitted using the JMAK equation. The conversion is found to be highly temperature-dependent, the extent of LDH platelets population (aLDH) being complete after 50 days, 21 days, 4 days, 3 days and 3 h for a temperature of 25 °C, 30 °C, 40 °C, 50 °C and 60 °C, respectively. Extrapolated values of the Avrami number n from the JMAK refinement, as well as from the Sharp-Hancock formalism (SEI 1) is spanning from 0.4 to 1.4, thus deviating strongly from the expected value of 1 for a layered system. The quality of the regression is acceptable, resulting in an estimated error in the value of k ranging roughly between 5% and 10%. However the shape of the conversion curves seems to be stretched at lower and compressed at higher temperatures respectively. Furthermore, at 30 °C, the curvature seems to adopt an intermediate step within a week. All together this underlines the strong effect of the temperature and indicates a complex scenario, mechanistically comprising at least two different processes that are superposing the LSH to LDH conversion. Besides a lowering of the pastes viscosity that allows for increased cation diffusion rates interfacial reaction zones build up a diffusion barrier for a continuous supply of Al3+ cations like the formation of a fringe layer that is surrounding aluminum rich domains. Aluminum cations will be consumed first for the conversion of small LSH platelets of that fringe layer before significant amounts are available for the conversion of bigger LSH layers in the bulk. The determination of the kinetic parameters k allows the extraction of the activation energy Ea according to the Arrhenius equation: Ea

k ¼ c:expRT where k, Ea, c, R and T correspond to the rate constant, the activation energy (kJ mol1), the pre-exponential factor, the perfect gas constant (R = 8.314 J mol1 K1) and the temperature (K), respectively. From JMAK data, an activation energy Ea = 144.98 kJ mol1 is

268

T. Stimpfling et al. / Journal of Colloid and Interface Science 462 (2016) 260–271

of alumino-silicate phosphate (SAPO-34) from a layered pre-phase requires as much as 160 kJ/mol [63]. These data support our interpretation that the process investigated here proceeds via platelets conversion through disintegration – rebuilding of the platelets. 3.5. Tentative interpretation of the mechanism Taking into account the XRD information and the observations revealed by TEM the role of the aluminum rich domains can be considered as a reservoir for aluminum cations needed to convert the initially formed LSH platelets, located in the zinc rich domains, into LDH layers; while their hairy layered corona that forms over time indicates that these eroding reservoirs are the location of a reaction zone for further layered material. From the different observations, a tentative scenario explaining the platelets transformation is pictured in Scheme 2. Although XRD suggests a complete LSH to LDH conversion the nature of these corona platelets remains a matter of speculation due their comparatively minor amount. However with regard to a local high aluminum cation concentration gibbsite and LDH are the most probable phases. For the latter tetrahedral zinc cations that have been liberated in the course of the LSH to LDH transformation would be available and would have to migrate into this zone of enriched aluminum. This scenario would be in line with the mechanism of substrate dissolution reported recently by Li et al. [32] and with the results regarding the activation energy. The progressive dissolution of LSH bulk implies a conversion occurring at the exterior of the particle that can be viewed as a peeling-off effect. Phase transformation may happen in a more

Fig. 8. (a) (1 1 0) diffraction peak deconvolution and (b) evolution of LDH population (aLDH) versus aging time at different indicated temperatures. Dashed lines correspond to refined plot by using general JMAK equation. ±10% value error bars are estimated.

Fig. 9. Arrhenius plot of ln k vs. 1/T for diffusion reaction model (Conversion LSH into LDH) determine by using general JMAK model. ±10% value error bars are estimated.

obtained (Fig. 9), that corresponds to the amount of energy needed to convert the LSH platelets into LDH in the presence of PES polymer and using Al(hydroxyl)acetate as reactants. Comparatively an activation energy of 27 kJ/mol is calculated for the diffusion of Li into c-Al(OH)3 [61,62], while the dissolution re-precipitation synthesis

Scheme 2. Simplified sketch of the LSH-LDH platelets transformation and LDH fringe formation. Tentative scenario: polymer (gray) intercalated LSH (blue) platelets transform into LDH (yellow) platelets (upper right) while LDH platelets grow as a fringe on the surface of aluminum rich particles (red, lower left), the latter serving as a reservoir for Al3+ cations (red dots). They diffuse to the LSH platelets in order to occupy octahedral lacunae, which liberates two Zn2+ cations (blue dots) per Al3+ uptake. Liberated as well as initially present zinc cations may form complexes with carboxylate groups from the polyester and react with aluminum cations at the surface of the aluminum rich particles forming a growing fringe layer (cation migration paths visualized as dotted lines). The topochemical LSH–LDH conversion as well as the diffusion rates of Al3+ and Zn2+, the latter being impacted by the diffusion barrier of a growing fringe layer and the complexing carboxlate groups of the polymer phase respectively, are kinetically reflected in the nonlinear conversion profile as a function of temperature. The topochemical LSH–LDH conversion is the rate determining process relative to the diffusion rates of Al3+ and Zn2+, the latter being impacted by the diffusion barrier of a growing fringe layer and the complexing carboxylate groups of the polymer phase, respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

T. Stimpfling et al. / Journal of Colloid and Interface Science 462 (2016) 260–271

complex way than initially thought as elegantly demonstrated by the sequential two separated delayed reaction in the case of LDH formation [18]. Taking into account the anisotropy and the possible topological front-reaction, a conversion may arise normal or parallel to the layers stacking. Conversion onto the platelets surface will probably result into blocking the reaction by encapsulating LSH inside, and forming a LDH shell that would be repulsive to Al3+ ingress. Indeed in the case of Mg:Al composition, large particles of magnesium hydroxide exposed to an aluminum hydroxycarbonate (AHCC) suspension demonstrate rapid hydrotalcite formation resulting in an occlusive coating [64]. Nucleation and growth of mixed crystal was occurring directly on the magnesium hydroxide surface, and was found to be rapid when mixing the separate suspension and to form inevitably magnesium aluminum double hydroxide material structurally similar to the mineral hydrotalcite [65]. The authors speculate that the magnesium hydroxide crystal structure may serve as a template for nucleation as a-parameters of Mg(OH)2 and LDH are nearly identical. In fact they suggested on the basis of acid neutralization experiments that brucite particles are covered and protected by aluminum hydroxide [66]. Interestingly the prevalent orientation of platelets normal to the aluminum reservoir surface (see after 120 days in Fig. 7c) appears to be the optimum geometry for an inter-diffusive reaction zone of Al3+ and Zn2+ ions and resembles the topochemical oxidative intercalation mechanism of brucite into LDH platelets which was recently described by Ma et al. [41] to occur along the basal direction within the sheets rather than across stacked layers. In the present case, the conversion proceeds at the interface, resulting in an observed hairy fringe on TEM pictures, thus implying that platelets are growing normal to the surface of the Al-rich domains rather than lying/stacking onto it. This intrinsic limit in the completeness of the conversion is demonstrated by the persistence of Al species not integrated within the LDH framework. It is interesting to note that for long aging times where the generation of LDH platelets is supposedly complete, small islands of Al-rich composition still remain (Fig. 7a). As Al3+ cations were 3+ introduced in a stoichiometric amount to reach a Zn2+ 0.66Al0.33 composition which is corroborated by the found cell parameter a = 0.304 nm (vide supra), the persistence of Al3+-zones must be concomitant with zones comprising other Zn compounds. We see the following possible explanations: (i) not all zinc based hydroxide material may exist in a sufficiently crystallized state giving rise to XRD diffraction, (ii) the aluminum rich phase may still contain zinc compound and finally (iii) the polymer matrix phase contains zinc cations, possibly complexed by carboxylate groups (SEI 2).

269

Fig. 10. Structural evolution from (a) Zn(OH)2 to (b) Zn2Al-LDH material after 48 h at 40 °C in PVP medium and using Al(Ac)2OH reactant.

zinc cations. Tethered hydrophobic polymer chains would thus stabilize the platelets toward the attack of hydrated aluminum cations. An indirect hint that PES might be involved in the conversion reaction stems from the LDH product that does not contain acetate as counter ion. TEM EDX analysis revealed the presence of zinc cations in the polyester rich bulk phase, an indication that bidirectional cation transport through the bulk phase, liberated former tetrahedral zinc cations toward the aluminum reservoir versus aluminum cations toward the LSH platelets, might proceed through a cascade of reversible complexation with carboxylate groups of the polyester. Similar to the polyester bearing system the process using PVP as polymer appears to be complete only looking at the crystallinity of the obtained material. However from 27Al MAS solid state NMR, it is obvious that the initial Al3+ source persists even after prolonged aging, thus underlining the presence of amorphous constituents within the water-soluble polymer (SEI 4). When changing Zn2+ for Co2+ cations, both LSH and LDH platelets are formed using the polyol route, and the XRD patterns displayed in Fig. 11 are consistent with the literature [48,49]. In the presence of PVP, LDH platelets are not directly formed but LSH. Aluminum(hydroxyl)acetate does not react immediately with Co

3.6. Extension of the lamellar transformation Finally, such a transformation was tentatively extended to other reactants, polymer and cation composition. In this idea, poly(vinyl) pyrrolidone (PVP) was selected since it is of importance in applications as a binder or film forming agent to produce porous polymer electrolytes by solvent evaporation [67]. Taking PVP as watersoluble polymer and the same acetate reactants as before, well crystallized Zn(OH)2 in place of LSH was obtained that converts into LDH platelets after 48 h at 40 °C (Fig. 10). In spite of a different structural path, the kinetics appear to be close to what was formerly observed with PES and LSH (from Fig. 8c, a  0,9 after 48 h). The precise role of the polymer is not fully understood yet, but it is crucial for the conversion route since without PVP, LDH platelets form directly at the end of the addition when isolated from the mother liquid and no intermediate phase is observed. It is reasonable to assume that LSH acetate undergoes a ligand exchange with the carboxylate groups of the polymer coordinating the tetrahedral

Fig. 11. Typical XRD pattern for Co-based (a) LSH and (b) LDH materials. The dashed lines materializing mostly the intralayer diffraction peak difference between LSH and LDH sheets as identified in Fig. 1.

270

T. Stimpfling et al. / Journal of Colloid and Interface Science 462 (2016) 260–271

Fig. 12. Typical DRX pattern of the structural transformation after 2 days from LSHCo to LDH-Co2Al using Al(hydroxyl)acetate and PVP as medium at 40 °C and the associated color change. The dashed lines materializing mostly the intralayer diffraction peak difference between LSH and LDH sheets as identified in Fig. 1. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

(acetate) but is kept as a Al3+ cation reservoir, and a shift of the diffraction line (110) is progressively observed over time and completed after 48 h at 40 °C (Fig. 12). This is also accompanied by color change from blue to brown–pink, consistent with the color change between the neat LSH and LDH platelets (Fig. 11). A similar color shift accompanies the LSH to LDH platelet conversion for the Co:Al cation composition in the presence of PES or PVP, respectively. The structural transformation over 5 days ageing yields very ill-defined in-plane cation organization (SEI 4 and 5, respectively). A peak around 2h = 9° is probably due to acetate-tethered platelets as observed in Fig. 11. It is worth noticing that using well-crystallized LSH platelets prepared without any polymer present (Fig. 1b), the conversion into LDH is not observed after the addition of an aluminum source and aging at 80 °C for extended times (not shown). This suggests that the platelets transformation requires well-exposed LSH platelets (interfacial ‘‘reaction zone”) and / or is entropically favoured, e.g. via acetate liberation as it is the case of polymer intercalated layers obtained by the in situ route. As mentioned in the introduction, LDH platelets formation is ubiquitous as demonstrated by earlier reports on hydrotalcite like compounds formed between an Al3+ source and solvated M2+ cations in aqueous medium [29–32], thus possibly yielding oriented and/or porous hydrotalcite like thin films [68]. Uncovered here in the course of approaches to novel polymer nanocomposite systems, the described structural transformation illustrates possibilities to tune between LSH and LDH platelets. Having in mind the chemical versatility of LSH that allows for multiple cationic compositions such as zinc copper hydroxyl acetate [69] or zinc nickel hydroxyl acetate [27] this intriguing ‘‘nano-process” opens up a possible new route for layered hydroxide based hybrid systems [70]. 4. Conclusions The structural transformation from LSH to LDH in the presence of an Al reservoir and a water dispersible polymer is here perused mostly through XRD, but also TEM and solid state 27Al NMR and using different chemical reactants and polymers. The salient fact is the ability of LSH to convert into LDH under certain conditions. The hitherto non avowed structural transformation is demonstrated through compelling evidence from the temperature dependent shift of the phase specific Bragg peak (1 1 0) positions of the intra layer ordering.

The first order topochemical reaction, featuring activation energy of about 140 kJ mol1, can be reconciled with a dissolution re-precipitation process rather than ingress of Al3+ cations into the lacunae of the LSH structure. The inward evolution of the restructuring involves the dissolution of two tetrahedral bound Zn2+ cations per octahedral void that is occupied by an Al3+ cation. The Avrami–Erofe’ev and more specifically the Johnson–Mhel–Av rami–Kolmogorov (JMAK) solid state kinetics models applied over the investigated temperature range reveal two different processes associated with different LSH–LDH platelets populations as visualized by TEM analysis. While neat acetate-LSH does not transform into LDH, the precise role of the polymer regarding anionic groups, architecture and amphoteric nature deserves further investigation. However by comparing different systems with regard to the used polymer and cation source two important structural features have been identified as crucial factors to enable this transformation within the studied time scale: well separated individual LSH platelets and a sufficient diffusivity for incoming Al3+ and outgoing Zn2+ cations between the aluminum reservoir and the reactive sites respectively. Uncovered by serendipity, the structural platelets conversion seems to inherit generalizability to be conceivably applied to further systems of layered inorganic materials and polymer as well as for other purposes than the design of polymer based composites. Acknowledgments We acknowledge SOLEIL for provision of synchrotron radiation facilities and we would like to thank Javier Perez for assistance in using beamline SWING. T.S. acknowledges the support from BASF Coatings GmbH. The authors thank Dr. Frechen (BASF SE) for cryo-TEM analyses. Appendix A. Supplementary material Sharp-Hancock method applied for a tentative determination of the LSH to LDH conversion at different temperatures. Summary of solid-state kinetic parameters for LSH to LDH conversion, and parameters n and k extracted from the general JMAK equation, and n and ln k from Sharp-Hancock plots. Refinement of the complete extent LDH (aLDH = 1) depending to the temperature. TEM of the evolution between Zn acetate and Al(hydroxyacetate) in PES after 120 h of time storage at room temperature. Extension of the structural transformation using PVP medium and Al(Ac)2OH showing the evolution by 27Al MAS NMR spectra (corresponding to platelets evolution shown in Fig. 10) and using PES medium. DRX of the structural transformation LSH Co in LDH with either Al(hydroxyl)acetate source in PES or AlCl3 source in PVP. The conversion CuLSH to Cu2Al LDH in PVP is also provided. Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jcis.2015.10.010. References [1] F. Bergaya, G. Lagaly (Eds.), Handbook of Clay Science, Part A Fundamentals, second ed., ISBN 978-0-08-098258-8. [2] F. Leroux, Recent Patent in Nanotechnology, 2012, vol. 6, pp. 157–249 (references therein). [3] U. Costantino, F. Leroux, M. Nocchetti, C. Mousty, LDH in physical, chemical, biochemical and life sciences, in: F. Bergaya, G. Lagaly (Eds.), Handbook of Clay Science, Part B. Techniques and Applications, Developments in Clay Science, vol. 5, second ed., Elsevier, 2013, pp. 765–791 (ISBN 978-0-08-098258-8 chapter 6). [4] J.B. Liang, R.Z. Ma, N. Iyi, Y. Ebina, K. Takada, T. Sasaki, Chem. Mater. 22 (2010) 371. [5] R. Ma, M. Osada, L.F. Hu, T. Sasaki, Chem. Mater. 22 (2010) 6341. [6] R.Z. Ma, T. Sasaki, Adv. Mater. 22 (2010) 5082. [7] M.F. Shao, M. Wei, D.G. Evans, X. Duan, Chem. A Eur. J. 19 (2013) 4100. [8] R. Ma, J.B. Liang, X.H. Liu, T. Sasaki, J. Am. Chem. Soc. 134 (2014) 19915.

T. Stimpfling et al. / Journal of Colloid and Interface Science 462 (2016) 260–271 [9] J.L. Gunjakar, I.Y. Kim, J.M. Lee, Y.K. Jo, S.J. Hwang, J. Phys. Chem. C 118 (118) (2014) 3847. [10] S. Cho, J. Kwag, S. Jeong, Y. Baek, S. Kim, Chem. Mater. 25 (2013) 1071. [11] R. Ma, X.H. Liu, J.B. Liang, Y. Bando, T. Sasaki, Adv. Mater. 26 (2014) 4173. [12] J.L. Gunjakar, T.W. Kim, H.N. Kim, I.Y. Kim, S.J. Hwang, J. Am. Chem. Soc. 133 (2011) 14998. [13] J.W. Zhao, J. Chen, S.M. Xu, M.F. Shao, Q. Zhang, F. Fei, J. Ma, M. Wei, D.G. Evans, X. Duan, Adv. Funct. Mater. 24 (2014) 2938. [14] Y. Liu, N. Wang, J.H. Pan, F. Steinbach, J. Caro, J. Am. Chem. Soc. http://dx.doi. org/10.1021/ja507408s. [15] M. Louër, D. Louër, D. Grandjean, Acta Cryst. B 29 (1973) 1696. [16] W. Stahlin, H.R. Oswald, Acta Cryst. B 26 (1970) 860. [17] S.J. Mills, A.G. Christy, J.M. Genin, T. Kameda, F. Colombo, Miner. Mag. 76 (2012) 1289. [18] Y. Yang, X. Zhao, Y. Zhu, F. Zhang, Chem. Mater. 24 (2012) 81. [19] J.W. Boclair, P.S. Braterman, Chem. Mater. 11 (1999) 298. [20] J.W. Boclair, P.S. Braterman, J.P. Jiang, S.W. Lou, F. Yarberry, Chem. Mater. 11 (1999) 303. [21] H. Roussel, V. Briois, E. Elkaim, A. de Roy, J.P. Besse, J.P. Jolivet, Chem. Mater. 13 (2001) 329. [22] T. Sato, H. Fujita, T. Endo, M. Shimada, React. Solids 5 (1988) 219. [23] T. Stanimirova, G. Kirov, Appl. Clay. Sci. 22 (2003) 295. [24] Z.P. Xu, G.Q. Lu, Chem. Mater. 17 (2005) 1055. [25] J.H. Choy, Y.M. Kwon, K.S. Han, S.W. Song, S.H. Chang, Mater. Lett. 34 (1998) 356. [26] E. Kandare, J.M. Hossenlopp, J. Phys. Chem. B 109 (2005) 8469. [27] R. Rojas, C. Barriga, M.A. Ulibarri, V. Rives, J. Solid State Chem. 177 (2004) 3392. [28] O. Palamarciuc, E. Delahaye, P. Rabu, G. Rogez, New J. Chem. 38 (2014) 2016. [29] J.L. Paulhiac, O. Clause, J. Am. Chem. Soc. 115 (1993) 11602. [30] J.B. d’Espinose de la Caillerie, M. Kermarec, O. Clause, J. Am. Chem. Soc. 117 (1995) 11471. [31] S.N. Towle, J.R. Bargar, G.E. Brown Jr., G.A. Parks, J. Colloid Interface Sci. 187 (1997) 62. [32] W. Li, K.J.T. Livi, W. Xu, M.G. Siebecker, Y. Wang, B.L. Phillips, D.L. Sparks, Environ. Sci. Technol. 46 (2012) 11670. [33] A. Miyasaki, M. Numata, M. Etou, K. Yonezu, I. Balint, T. Yokoyama, T. Colloid Surf. A 420 (2013) 115. [34] X. Liu, R. Ma, Y. Bando, T. Sasaki, Angew. Chem. Int. Ed. 49 (2010) 8253. [35] V.Z. Kohlschütter, Anorg. Allg. Chem. 405 (1919) 1. [36] W. Feitknecht, Fortschr. Chem. Forschung 2 (1952) 670. [37] V.V. Boldyrev, React. Solids 8 (1990) 231. [38] R. David, H. Kabbour, D. Filimonov, M. Huve, A. Pautrat, O. Mentre, Angew. Chem. 53 (2014) 13365.

271

[39] J.F. Wang, G.N. Zhang, L.J. Ren, L.P. Kang, Z.P. Hao, Z.B. Lei, Z.H. Liu, Cryst. Growth Design 14 (2014) 5626. [40] D. Montasserasadi, D. Mohanty, A. Huq, L. Heroux, E.A. Payzant, J.B. Wiley, Inorg. Chem. 53 (2014) 1773. [41] M. Ogawa, M. Hiramine, Cryst. Growth Design 14 (2014) 1516. [42] R. Ma, J. Liang, X. Liu, T. Sasaki, J. Am. Chem. Soc. 134 (2012) 19915. [43] Z.P. Xu, H.C. Zeng, Chem. Mater. 12 (2000) 2597. [44] P. Vialat, C. Mousty, C. Taviot-Guého, G. Renaudin, H. Martinez, J.C. Dupin, E. Elkaim, F. Leroux, Adv. Funct. Mater. 24 (2014) 4831. [45] J.P. Thiel, C.K. Chiang, K.R. Poeppelmeier, Chem. Mater. 5 (1993) 297. [46] A.V. Besserguenev, A.M. Fogg, R.J. Francis, S.J. Price, D. O’Hare, V.P. Isupov, B.P. Tolochko, Chem. Mater. 9 (1997) 241. [47] S. Komarneni, N. Kozai, R. Roy, J. Mater. Chem. 8 (1998) 1329. [48] L. Poul, N. Jouini, F. Fievet, Chem. Mater. 12 (2000) 3123. [49] V. Prevot, C. Forano, J.P. Besse, Chem. Mater. 17 (2005) 6695. [50] G.V. Manohara, P.V. Kamath, W. Milius, J. Solid State Chem. 196 (2012) 356. [51] N. Iyi, T. Sasaki, J. Colloid Interface Sci. 322 (2008) 237. [52] A.L. Troutier-Thuilliez, H. Hintze-Bruening, C. Taviot-Guého, V. Verney, F. Leroux, Soft Matter 7 (2011) 4242. [53] E. Hosono, S. Fujihara, T. Kimura, H. Imai, J. Colloid Interface Sci. 272 (2004) 391. [54] A. Kasai, S. Fujihara, Inorg. Chem. 45 (2006) 415. [55] L.Y. Zhu, S.P. Su, J.M. Hossenlopp, Polym. Adv. Technol. 23 (2012) 171. [56] S. Majoni, J.M. Hossenlopp, Polym. Degrad. Stab. 95 (2010) 1593. [57] F. Geng, R. Ma, Y. Ebina, Y. Yamauchi, N. Miyamoto, T. Sasaki, J. Am. Chem. Soc. 136 (2014) 5491. [58] P. Vialat, C. Taviot-Guého, G. Villemure, F. Leroux, C. Mousty, Electrochem. Acta 107 (2013) 599. [59] R. Marangoni, M. Bouhent, C. Taviot-Gueho, F. Wypych, F. Leroux, J. Colloid Interface Sci. 333 (2009) 120. [60] A. Khawam, D.R. Flanagan, J. Phys. Chem. B 110 (2006) 17315. [61] D. O’Hare, J.S.O. Evans, A. Fogg, S. O’Brien, Polyhedron 19 (2000) 297. [62] G.R. Williams, N.H. Rees, D. O’Hare, Solid State Sci. 11 (2009) 1229. [63] O.B. Vistad, D.E. Akporiaye, K.P. Lillerud, J. Phys. Chem. B 105 (2001) 12437. [64] W.J. Mc Laughlin, J.L. White, S.L. Hem, J. Colloid Interface Sci. 165 (1994) 41. [65] R.K. Vanderlaan, J.L. White, S.L. Hem, J. Pharm. Sci. 71 (1982) 780. [66] R.K. Vanderlaan, J.L. White, S.L. Hem, J. Pharm. Sci. 68 (1979) 1498. [67] F. Leroux, K. Chabrol, S. Morlat Thérias, J.L. Gardette, A. de Roy, Eur. J. Inorg. Chem. 32 (2012) 5332. [68] Y.F. Gao, M. Nagai, Y. Masuda, F. Sato, W.S. Seo, K. Koumoto, Langmuir 22 (2006) 3521. [69] C. Machingauta, J.M. Hossenlopp, J. Colloid Interface Sci. 411 (2013) 129. [70] G. Rogez, C. Massobrio, P. Rabu, M. Drillon, Chem. Soc. Rev. 40 (2011) 1031.