Kinetics and Mechanism of Calcium Hydroxide Conversion into

May 5, 2016 - ... as opposed to stable calcite which forms in untransformed Ca(OH)2 samples. Similar effects are obtained when a commercial nanolime p...
5 downloads 11 Views 4MB Size
Download PDF
Article pubs.acs.org/Langmuir

Kinetics and Mechanism of Calcium Hydroxide Conversion into Calcium Alkoxides: Implications in Heritage Conservation Using Nanolimes Carlos Rodriguez-Navarro,* Irene Vettori, and Encarnacion Ruiz-Agudo Department of Mineralogy and Petrology, University of Granada, Fuentenueva s/n, 18002 Granada, Spain ABSTRACT: Nanolimes are alcohol dispersions of Ca(OH)2 nanoparticles used in the conservation of cultural heritage. Although it was believed that Ca(OH)2 particles were inert when dispersed in short-chain alcohols, it has been recently shown that they can undergo transformation into calcium alkoxides. Little is known, however, about the mechanism and kinetics of such a phase transformation as well as its effect on the performance of nanolimes. Here we show that Ca(OH)2 particles formed after lime slaking react with ethanol and isopropanol and partially transform (fractional conversion, α up to 0.08) into calcium ethoxide and isopropoxide, respectively. The transformation shows Arrhenius behavior, with apparent activation energy Ea of 29 ± 4 and 37 ± 6 kJ mol−1 for Caethoxide and Ca-isopropoxide conversion, respectively. High resolution transmission electron microscopy analyses of reactant and product phases show that the alkoxides replace the crystalline structure of Ca(OH)2 along specific [hkl] directions, preserving the external hexagonal (platelike) morphology of the parent phase. Textural and kinetic results reveal that this pseudomorphic replacement involves a 3D diffusion-controlled deceleratory advancement of the reaction front. The results are consistent with an interface-coupled dissolution−precipitation replacement mechanism. Analysis of the carbonation of Ca(OH)2 particles with different degree of conversion into Ca-ethoxide (α up to 0.08) and Ca-isopropoxide (α up to 0.04) exposed to air (20 °C, 80% relative humidity) reveals that Ca-alkoxides significantly reduce the rate of transformation into cementing CaCO3 and induce the formation of metastable vaterite, as opposed to stable calcite which forms in untransformed Ca(OH)2 samples. Similar effects are obtained when a commercial nanolime partially transformed into Ca-ethoxide is subjected to carbonation. Such effects may hamper/delay the strengthening or consolidation effects of nanolimes, thus having important implications in the conservation of cultural heritage.

■

precursors10,17 and anion-exchange resins.18 The resulting portlandite nanocrystals (ca. 30−300 nm in size) were dispersed in short-chain aliphatic alcohols such as ethanol, propanol, or isopropanol, forming very stable colloidal dispersions that displayed high penetration efficiency.8 These features enabled these so-called nanolimes to induce a deep consolidation once applied onto porous supports (e.g., stone) while preventing surface accumulation leading to undesired glazing. Until recently, it was believed that once dispersed in alcohol, Ca(OH)2 nanoparticles were inert; that is, no phase transformation occurred during storage or/and handling of the dispersion prior to application. This common belief appeared to be justified by the assumed insolubility of Ca(OH)2 in alcohol. Nonetheless, it has been reported that portlandite is not strictly insoluble in alcohol: for instance, its solubility in methanol is ∼0.1 g/L.19 We recently showed that, in fact, upon contact with

INTRODUCTION Nanolimes are alcohol dispersions of Ca(OH)2 (portlandite) nanoparticles that have emerged as an efficient and compatible conservation material for the consolidation of degraded stones, plasters, and mural paintings1−8 as well as archeological and paleontological bones.9 They have also found applications in the conservation of other cultural heritage materials such as paper,10 canvas,11 and wood objects12 subjected to damage associated with acidification. While in the former cases the nanoparticles consolidate and strengthen the treated porous material (stone, plaster, or mortar) following the formation of cementing CaCO3 after carbonation via the reaction Ca(OH)2 + CO2 = CaCO3 + H2O, in the latter cases, the nanoparticles act as a pH buffer (i.e., they increase alkalinity), thereby preventing further chemical weathering. In their pioneering work, Giorgi et al.13 obtained stable alcohol dispersions of Ca(OH)2 particles produced after lime slaking according to the reaction CaO + H2O = Ca(OH)2. Subsequently, colloidal Ca(OH)2 nanoparticles were synthesized in diols,14 aqueous solutions with or without additives,1,15 and water-in oil microemulsions16 as well as using alkoxide © 2016 American Chemical Society

Received: March 18, 2016 Revised: May 5, 2016 Published: May 5, 2016 5183

DOI: 10.1021/acs.langmuir.6b01065 Langmuir 2016, 32, 5183−5194

Article

Langmuir

morphology with size (maximum length) ranging from 30 to 400 nm and mean size of ∼100 nm (TEM observations). These primary crystals tend to aggregate,21 forming larger secondary particles 200 nm up to 1.5 μm in size. N2 sorption measurements show that the average BET surface area of oven-dried FG putty is 8 ± 2 m2/g. Alcohol Dispersions and Thermal Treatment. Ethanol (Panreac, puriss.) and isopropanol (Panreac, puriss.) dispersions of Ca(OH)2 particles were prepared. The solid content was 5 g/L, as it is common in commercial nanolimes.22 Dispersion aliquots (100 mL) were placed in plastic bottles (150 mL), which were tightly capped to avoid solvent evaporation. The bottles were subjected to sonication for 5 min and placed in separate ovens set at 40, 60, and 80 °C. At predetermined time intervals (7 days elapsed time) dispersion aliquots were collected and subjected to drying in an air-ventilated oven at 80 °C for 72 h. Dry solids were stored in airtight plastic vials prior to further analysis. Sample collection was discontinued after two months because TG analyses (see below) showed no further conversion into Ca-alkoxide after ca. 6 weeks. In order to prevent the possible amorphization of newly formed Ca-alkoxides due to desolvation during oven-drying,23−25 selected dispersion samples were subjected to vacuum-filtering (using Millipore membrane filters, ϕ = 0.2 μm) and subsequent drying in air at room T, prior to XRD analysis. Note that in addition to being the most common alcohols used for the dispersion of nanolimes, we used ethanol and isopropanol in order to identify whether of not the type of alcohol (i.e., alkyl chain length) played a role in the conversion kinetics and carbonation performance of dispersed Ca(OH)2 particles. Analysis of Reactant and Product Phases. The mineralogy of solids was determined by XRD on a PANanalytical XPert Pro with Ni filter. Measurement parameters were Cu Kα radiation λ = 1.5405 Å, 45 kV, 40 mA, 4°−70° 2θ exploration range, steps of 0.001° 2θ, and goniometer speed of 0.01° 2θ s−1. Powders were deposited in zerobackground Si sample holders. Mineral phases were identified by comparison with JCPDS powder spectra (Joint Committee on Powder Diffraction Standards). TG and differential scanning calorimetry (DSC) analyses were performed simultaneously on a Mettler-Toledo TGA/DSC1 coupled to an FTIR (ThermoFisher Nicolet IS10) for evolved gas analysis. Samples of ∼50 mg were placed in Al crucibles and analyzed in flowing N2 (50 mL/min) at a heating rate of 20 °C/min (25−950 °C temperature interval). Additional compositional and microstructural features of Ca(OH)2 particles were determined by means of TEM (Titan, 300 kV acceleration voltage) and FTIR (Nicolet 20SXB with a resolution of 0.4 cm−1; KBr pellets were used) analyses. The (apparent) static contact angle, θ, of unreacted and reacted Ca(OH)2 particles was measured using the sessile drop method. Alcohol dispersions were deposited onto glass slides and let to dry in air at room T for 1 h. Subsequently, 10 μL droplets of Milli-Q water were deposited onto the dried film of Ca(OH)2 particles, and images were immediately collected. Following measurement of the height, h, and base, d, of the droplets, θ was calculated according to26

alcohol (e.g., ethanol or isopropanol) Ca(OH)2 nanoparticles partially transformed into Ca-alkoxides via the reaction Ca(OH)2 + 2ROH ⇄ Ca(OR)2 + 2H2O (where R is an alkyl group).6 We also showed that the release of alcohol during carbonation of Ca(OH)2 nanoparticles partially transformed into Ca-alkoxides, induced the formation of vaterite with a lower consolidation capacity than calcite, which is the stable phase formed upon carbonation of pure Ca(OH)2. Furthermore, we reported that the presence of Ca-alkoxides appeared to increase the rate of nanolime carbonation (and its yield). However, this latter effect could not be easily gauged because three different Ca(OH)2 (nano)particle dispersions with significant differences in particle size, surface area (i.e., reactivity), and degree of conversion into Ca-alkoxides were used in our previous study.6 While it is now known that Caalkoxides can form in nanolime alcohol dispersions, little is known, however, about the mechanism and kinetics of Ca(OH)2 to Ca-alkoxide conversion, and a detailed analysis of the effects of such a conversion on the kinetics of carbonation is lacking. It is our aim to shed light on the mechanism and kinetics of the transformation of Ca(OH)2 particles into Ca-alkoxides once dispersed in short-chain alcohols and disclose what is the exact effect such a conversion brings about on the kinetics of carbonation. In particular, here we re-evaluate our previous results that suggested an enhancement of the carbonation rate and yield of nanolimes subjected to transformation into Caalkoxides, observing that this is not necessarily the case. For these tasks we characterized the mineralogical/compositional and microstructural features of Ca(OH)2 particles dispersed in ethanol and isopropanol (i.e., the two most common alcohols used for the preparation of nanolime dispersions),4 once subjected to storage for different periods of time at three T (40, 60, and 80 °C). Using thermogravimetry (TG), we quantified the t-dependent degree of conversion into Ca-alkoxide and obtained kinetic parameters for such a conversion (reaction model, reaction rates, apparent activation energy, Ea, and preexponential factor, A). The kinetic study in combination with X-ray diffraction (XRD), Fourier transform infrared spectroscopy (FTIR), and transmission electron microscopy (TEM) analyses of reactant and product phases enabled us to propose a mechanism for the conversion of Ca(OH)2 particles into Caalkoxides. We also quantified the amount (and phase) of CaCO3 formed over time during carbonation of Ca(OH)2 particles with different degree of conversion into Ca-alkoxides. This enabled us to gather quantitative information regarding the effects of such a conversion on carbonation rate and yield. We also tested the effects that a partial conversion into Caalkoxide of a commercial nanolime produces on its carbonation rate, CaCO3 yield, and polymorph selection. Finally, we discuss the implications such a conversion have in the conservation of cultural heritage using nanolimes.

■

θ h = tan−1 2 0.5d

(1)

Kinetic Analysis. TG results were used to determine the evolution of the fractional conversion, α (α = Xt/X0, where X0 and Xt are the molar fraction of Ca(OH)2 at time 0 and at time t, respectively) of Ca(OH)2 into Ca-alkoxide(s) vs t for the tested T-runs (40, 60, and 80 °C). Experimental α values were used to determine the apparent rate constant k(T), considering the relationship

MATERIALS AND METHODS

Ca(OH)2 Particles. As starting material we choose to use a slaked lime putty (FG) whose physical−chemical, mineralogical, and microstructural properties have been fully characterized previously.6,20 This enabled us to have a sufficient amount of well-characterized material for all the different experimental runs. In brief, this slaked lime is made up of Ca(OH)2 particles with small amounts of calcite (CaCO3) (∼4.5 wt %, according to TG analysis) formed due to early carbonation during sample storage and preparation (i.e., drying of the lime putty). Ca(OH)2 particles show a mean crystallite size of 96 ± 11 nm (XRD results) and display a predominantly platelike hexagonal

dα = f (α)k(T ) dt

(2)

where f(α) is the conversion function describing the reaction model (see a full list of reaction models in refs 27 and 28). Replacing k(T) with the Arrhenius equation yields 5184

DOI: 10.1021/acs.langmuir.6b01065 Langmuir 2016, 32, 5183−5194

Article

Langmuir

Figure 1. FTIR and TG/DSC analysis of the conversion of Ca(OH)2 into Ca-alkoxides. FTIR spectra (a, c) and TG/DSC traces (b, d) of portlandite particles subjected to transformation into Ca-ethoxide (a, b) and Ca-isopropoxide (c, d) at different T (40, 60, and 80 °C) for two months. The colored shaded bar in the FTIR spectra marks the position of the C−H stretching mode of CH3 and CH2 groups characteristic of alkoxides. The brownish, blue, and green shaded areas in the TG/DSC graphs mark the T-interval for Ca-alkoxide decomposition, portlandite dehydroxylation, and CaCO3 thermal decomposition, respectively.

⎛ −E ⎞ dα = A exp⎜ a ⎟f (α) ⎝ RT ⎠ dt

± 2 °C and 80 ± 5% relative humidity (RH). The container was not airtight in order to allow a small but continuous flux of air (pCO2 ∼ 10−3.5 atm) for carbonation to take place. A high RH was selected in order to speed up the carbonation process.6,29 Note, however, that the specific crystalline phases formed after Ca(OH)2 carbonation in air are dependent on RH (and T).3 Samples were analyzed by XRD at predetermined time intervals (up to 39 days), and the degree of transformation of Ca(OH)2 into CaCO3 and the ratio (mass fraction) of calcite (Cc) to vaterite (V) in carbonated Ca(OH)2 were calculated following the procedure described elsewhere.6 Effects of the Partial Conversion of a Commercial Nanolime into Ca-Alkoxide. A commercial nanolime (CaLoSil E-25; IBZSalzchemie, Freiberg, Germany) was obtained. CaLoSil E-25 is an ethanol dispersion of colloidal Ca(OH)2 nanoparticles (25 wt % solid content). Nanoparticles are prepared following hydrolysis of a Caalkoxide precursor and subsequent dispersion into ethanol.17 The dispersions are made up of nearly monodisperse platelike Ca(OH)2 nanoparticles ∼100−200 nm in size with a BET surface area of 31.5 ± 0.5 m2/g. One batch of the nanolime ethanol dispersion was subjected to conversion into Ca-ethoxide. For this task, an airtight plastic bottle (volume 100 mL) was placed for up to 2 weeks into an oven set at 60 °C. Following its partial conversion into Ca-ethoxide (evaluated using

(3)

where A and Ea are the Arrhenius parameters and R is the gas constant. Integration of eq 2 gives

g (α) = k(T )t

(4)

where g(α) is the integrated form of the reaction model f(α). Selection of a particular reaction model best describing the experimental results was performed by linear fitting of g(α) vs t. This enabled us to obtain the corresponding rate constant k(T) for the three T of our experimental runs. The Arrhenius parameters were finally obtained using the logarithmic form of the Arrhenius equation ln k(T ) = ln A −

⎛ Ea ⎞ ⎜ ⎟ ⎝ RT ⎠

(5)

Carbonation of Partially Transformed Ca(OH)2 Particles. Alcohol dispersions of Ca(OH)2 particles (ca. 2 mL) with different degree of fractional conversion (up to the maximum achieved) into Ca-ethoxide (α from 0 to 0.08) and Ca-isopropoxide (α from 0 to 0.04) were deposited on glass slides and let to dry in air at room T for 60 min. Once dried, the samples were place in a plastic container at 18 5185

DOI: 10.1021/acs.langmuir.6b01065 Langmuir 2016, 32, 5183−5194

Article

Langmuir FTIR and TG-DSC), ethanol dispersions of this nanolime and of an untransformed nanolime (control) were deposited on glass supports, allowed to dry in air at room T for 1 h, and subsequently subjected to carbonation at 80 ± 5% RH (room T) following the experimental protocol outlined above. The carbonation degree and CaCO3 phase(s) vs time were evaluated using XRD (as indicated in the previous section).

earth alkoxides are typical solvated; hence, a more general formula should be Ca(OCH2CH3)2·nHOCH2CH3, with n = 2 (for reaction at T ≥ 60 °C) or n = 4 (for reaction at 40 °C).33 Note also that crystalline Ca(OCH2CH3)2·2HOCH2CH325 readily desolvates (e.g., within 72 h when placed in a desiccator at room T).23 It is thus expected that the oven-drying treatment to which the samples were subjected prior to analysis (72 h at 80 °C) resulted in the full desolvation of the alkoxide. Similarly, in the case of the isopropanol dispersions, the new phase was a calcium isopropoxide with formula Ca(OCH2CH2CH3)2. As in the previous case, this alkoxide was likely solvated prior to oven-drying. The most likely overall reactions taking place during the thermal decomposition of Ca-ethoxide and Ca-isopropoxide in N2 atmosphere are

■

RESULTS AND DISCUSSION Transformation of Ca(OH)2 into Ca-Alkoxides. FTIR spectra of all studied FG samples (slaked lime dispersed in alcohol) systematically showed a strong and sharp absorption band at 3641 cm−1 and a broad band centered at 3490 cm−1 corresponding to the OH stretching modes as well as a small band at 1655 cm−1 corresponding to the OH bending, all of them characteristic of crystalline Ca(OH)2 (Figures 1a and 1c). A broad band was also observed centered at ca. 1480 cm−1 along with a small sharp band at 877 cm−1, corresponding to the ν3 asymmetric stretching and the ν2 symmetric deformation of carbonate groups, respectively, consistent with the presence of small amounts of CaCO3 in the original FG slaked lime putty dispersion. The small doublet at 2500−2530 cm−1 corresponds to atmospheric CO2 (a band which could not be fully eliminated during background subtraction). Samples subjected to reaction in ethanol and isopropanol showed additional bands (triplet) at 2830−2960 cm−1 corresponding to the C−H stretching mode of CH3 and CH2 characteristic of alkoxides.25,30 The intensity of these absorption bands increased with reaction time and oven T and was systematically higher in samples dispersed in ethanol if compared to those dispersed in isopropanol (Figures 1a and 1c). TG/DSC analyses showed that control samples displayed a main weight loss in the T range 410−580 °C corresponding to the dehydroxylation of portlandite according to the reaction Ca(OH)2 = CaO + H2O. This event was endothermic as shown by the DSC trace (Figures 1b and 1d). In contrast, TG traces of samples subjected to reaction in alcohol showed an additional weight loss between 100 and 410 °C, also endothermic (which overlapped with the DSC trace of Ca(OH)2 dehydroxylation, thus yielding a split band). This weight loss is associated with the thermal decomposition of alkoxides.30,31 In addition, samples subjected to reaction in alcohol displayed a small exothermic peak at ∼350 °C, more marked in the samples reacted in isopropanol (Figures 1b and 1d). The origin of this exothermic peak is not clear. It could be hypothesized that it corresponds to the crystallization of Ca(OH)2 after decomposition of the Ca-alkoxides, which involves dehydration of the alcohol (see below) and the subsequent hydration of CaO formed after thermal decomposition of the Ca-alkoxides. Conversely, it is also possible that such an exothermic peak might correspond to the crystallization of an amorphous calcium carbonate (ACC) phase, which reportedly occurs at ∼330−350 °C.32 Such an amorphous phase may form after partial carbonation of Ca(OH)2 exposed to H2O and CO2 released during Ca-alkoxide thermal decomposition (see below). The magnitude of the weight lost in the 100−410 °C range was directly related to the duration and oven T at which each Ca(OH)2 sample was subjected to reaction with ethanol or isopropanol. These results confirm the new formation of Ca-alkoxides.6 In the case of ethanol dispersions, the new phase, a calcium ethoxide,30 formed according to the reaction Ca(OH)2 + 2HOCH2CH3 = Ca(OCH2CH3)2 + 2H2O. Note that alkaline-

Ca(OCH 2CH3)2 = CaO + 2C2H4 + H 2O

(6)

Ca(OCH 2CH 2CH3)2 = CaO + 3C2H4 + H 2O

(7)

In addition to these well-known dehydration reactions, we cannot rule out the possibility of alcohol decomposition via dehydrogenation (forming acetaldehyde), as well as steam reforming reactions (associated with H2O release during the dehydration reactions), resulting in the formation of CH4 (plus H2, CO, and CO2).34 This is consistent with FTIR analysis of evolved gases showing the release of both methane and ethylene, in addition to acetaldehyde, water, CO2, and CO (note that H2 could not be detected by FTIR). Overall, this decomposition route differs from that proposed by Dongare and Sinha31 for the case of Ca-ethoxide. The authors observed the release of ethanol and ethylene at 220 °C, followed by the decomposition of the remaining ethoxide (at 460 °C), resulting in a mixture of CaO and Ca(OH)2, and the release of ethylene and ethanol. The authors reported that this second stage was followed by the high-T (860 °C) decomposition of Ca(OH)2. We have found no evidence for the direct release of ethanol (or isopropanol) or the subsistence and dehydroxylation of Ca(OH)2 at T > 500 °C. XRD analyses of Ca(OH)2 particles with different degree of transformation into Ca-ethoxide or Ca-isopropoxide did not reveal the presence of any Bragg peaks corresponding to a crystalline alkoxide. This shows that the newly formed Caalkoxides were amorphous to X-rays (their amorphous nature was also confirmed by TEM analysis; see below). Apparently, amorphization resulted from desolvation during oven-drying of the reacted Ca(OH)2 alcohol dispersions. We attempted to bypass such a possible effect by vacuum filtering the reacted Ca(OH)2 dispersions at room T (see Materials and Methods section). No crystalline phase, other than portlandite (and trace amounts of calcite), could be identified in samples treated at 40 and 60 °C, where the fractional conversion into Ca-alkoxide was 75%.29 In air, at the low T of our tests, the possibility of carbonation progressing at a very slow rate via a solid-state diffusion mechanism is ruled out. Water is required for carbonation to take place because, as recently demonstrated, the carbonation of portlandite in air at STP conditions is a dissolution−precipitation process (i.e., it is not a solid state process).51 It takes place within a film of water molecules (saturated with respect to portlandite) at the surface of Ca(OH)2 particles.50 CO2 can dissolve within this film and transform into CO32− ions that can react with Ca2+ released from portlandite dissolution, finally leading to CaCO 3 precipitation (within such an aqueous film).51 Interestingly, it has been reported that the formation of alkoxides on a range of hydrophilic solid substrates (e.g., calcite, TiO2, and CaO) following (chemi)sorption of aliphatic alcohols made them hydrophobic.47,53−55 This is due to the hydrophobic character of the alkyl groups (methyl, ethyl, or isopropyl) present in the alkoxides.56 This effect has been confirmed by water vapor adsorption isotherms showing a significant reduction in adsorbed H2O onto alkoxide-modified hydrophilic substrates (e.g., TiO2).47 In addition, it has been reported that the dissolution rates of ionic crystals (e.g., calcite) is reduced following (chemi)sorption of aliphatic alcohols.57 The previous effects can help to explain why the formation of Ca-alkoxides on the surface of Ca(OH)2 particles reduces so drastically their carbonation rate. Our contact angle measurements show that upon partial conversion into Ca-alkoxide a film of Ca(OH)2 particles changes its contact angle from ∼10° up to ∼60°, which indicates that the particles are rendered partially hydrophobic. This should difficult multilayer H2O adsorption (at 80% RH), thus slowing the hydrolysis of the Ca-alkoxides as well as the dissolution of the Ca(OH)2 substrate and hampering the progress of the carbonation process. Furthermore, Ca-alkoxides seem to form an armor around the Ca(OH)2 nanoparticles, preventing the access of H2O (and dissolved CO2) to the unreacted portlandite core. To these effects, it has to be added the above-indicated passivating effect that the precipitation of CaCO3 on Ca(OH)2 crystals brings about. Overall, the combination of all the previous effects helps to explain the drastic reduction in carbonation rate and CaCO3 yield observed here upon partial transformation of Ca(OH)2 into Ca-alkoxides. Another relevant observation was the systematic formation of calcite upon carbonation of control samples (not subjected to Ca-alkoxide conversion) and the formation of vaterite in samples subjected Ca-alkoxide conversion either in ethanol or in isopropanol (Figures 4c and 4d). Note that vaterite was the only phase present after carbonation of the samples displaying the highest degree of conversion into Ca-alkoxide. It has been thoroughly reported that a number of organic molecules (e.g., alcohols) tend to favor the formation and kinetic stabilization of vaterite, preventing/delaying its transformation into stable calcite.58,59 However, our samples subjected to reaction in 5190

DOI: 10.1021/acs.langmuir.6b01065 Langmuir 2016, 32, 5183−5194

Article

Langmuir alcohol and conversion into Ca-alkoxide were dried prior to carbonation. So, no free alcohol was expected to be present during the conversion of portlandite into CaCO3. Nonetheless, Ca-alkoxides can undergo hydrolysis upon exposure to humidity (although limited under our experimental conditions: see above), thereby transforming into Ca(OH)2 and releasing the constituent alcohol. Such a released alcohol, once dissolved into the aqueous solution (film) needed for Ca(OH) 2 carbonation, would influence CaCO3 polymorph selection (i.e., vaterite) as we have reported elsewhere.6 Note that formation and kinetic stabilization of metastable vaterite has been shown to lead to a lower consolidation strengthening than stable calcite following treatment of porous stones with alcohol dispersions of Ca(OH)2 nanoparticles.6 Nonetheless, due to its metastable nature, vaterite should eventually convert into stable calcite. Indeed, it has been reported that in solution metastable CaCO3 precursors such as vaterite rapidly convert into stable calcite (within hours) via a dissolution−precipitation mechanism (see ref 59 and references therein). However, the kinetics of such a phase transformation in air have not been studied yet. Carbonation of a Commercial Nanolime. To test if the observed detrimental effects associated with the partial transformation into Ca-alkoxides (slow carbonation rate, limited CaCO3 yield and preferential formation of metastable vaterite) taking place in the case of the alcohol dispersions of slaked lime particles also occurred in the case of commercial nanolimes used in heritage conservation, the following study was performed. A commercial nanolime, CaLoSil, was partially transformed into Ca-ethoxide (α = 0.08) following storage for up to 2 weeks in an oven set at 60 °C (see FTIR spectra in Figure 5a and TG traces in Figure 5b). Upon exposure to humid air (80% RH, 20 °C), the control (fresh, untransformed commercial nanolime) readily carbonated, reaching a CaCO3 yield of ca. 100% within 48 h (Figure 6). In contrast, the nanolime partially converted into Ca-ethoxide required 6 days to reach a carbonation of ∼88%, and then on the carbonation rate was drastically reduced (Figure 6). Full carbonation was not achieved until 2 weeks. Furthermore, while carbonation of the former nanolime led to the formation of a mixture of calcite (75 ± 14 wt %) with minor amounts of vaterite (25 ± 14 wt %), the latter resulted in ∼100 wt % vaterite. These results are consistent with those reported by Rodriguez-Navarro et al.6 showing that the carbonation of a commercial nanolime displaying a high degree of conversion into Ca-isopropoxide took more than 10 days (under the same experimental conditions) and resulted in 100% vaterite. Overall, these results demonstrate that (i) high-surface area commercial nanolimes are apparently more reactive toward carbonation than the lowsurface area Ca(OH)2 particle dispersions prepared from slaked lime. It should be noted, however, that the surface area of the commercial nanolime is 4 times higher than that of the slaked lime Ca(OH)2 particles. This could explain the apparently higher carbonation rate of the former. However, normalization with respect to surface area shows that both the commercial nanolime and the Ca(OH)2 dispersion prepared using slaked lime carbonate at a similar rate (within error). In any case, the yield of the former is ∼100%, while that of the latter is ∼27%. Ultimately, their carbonation behavior will not be simply related to surface area but also to the size and shape of the particles, which in the case of the commercial nanolime enable the progress of the reaction to the core of Ca(OH) 2 nanoparticles until completion, as explained above. (ii) Conversion of commercial nanolimes into Ca-alkoxides reduces

Figure 5. Transformation of commercial nanolime into Ca-ethoxide. (a) FTIR spectra and (b) TG/DSC traces of CaLoSil nanolime subjected to transformation into Ca-ethoxide for 14 days in ethanol at 60 °C. The blue shaded bar in the FTIR spectra marks the position of the C−H stretching mode of CH3 and CH2 groups characteristic of alkoxides. The brownish, blue, and green shaded areas in the TG/DSC graphs mark the T-interval for Ca-alkoxide decomposition, portlandite dehydroxylation, and CaCO3 thermal decomposition, respectively.

their carbonation rate, although to a lower extent than in the case of portlandite nanoparticles obtained via lime slaking (likely because the smaller particle size and higher surface areaand reactivityof commercial nanolimes somehow counteracts the passivating effect of the Ca-alkoxides). (iii) Carbonation of commercial nanolimes partially converted into Ca-alkoxide favors the formation of vaterite, the less stable CaCO3 polymorph (as also occurs in the case of slaked lime nanoparticles partially transformed into Ca-alkoxides). It follows that in the case of commercial nanolimes the partial transformation of Ca(OH)2 nanoparticles into Ca-alkoxides is detrimental for their carbonation and, in turn, should be a handicap for their performance during conservation interventions. Henceforth, caution should be taken when storing nanolime alcohol dispersions in order to prevent/limit their (partial) conversion into Ca-alkoxides. In any case, our experimental results show that, ultimately, the Ca-alkoxides will eventually hydrolyze and convert into Ca(OH)2. This phase in turn will undergo full carbonation. It is 5191

DOI: 10.1021/acs.langmuir.6b01065 Langmuir 2016, 32, 5183−5194

Langmuir

■

Article

AUTHOR INFORMATION

Corresponding Author

*Tel +34 958 246616; Fax +34 958 243368; e-mail carlosrn@ ugr.es (C.R.-N.). Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work has been financially supported by the Spanish government (Contracts CGL2012-35992 and CGL201570642-R) and the Junta de Andalucı ́a Research Group RNM179 and project P11-RNM-7550. E.R.-A. acknowledges the receipt of a Ramón y Cajal grant. TG/DSC and TEM analyses were performed at the Centro de Instrumentación Cientı ́fica of the Universidad de Granada. We thank K. Kudlacz for his help with the calibration curves for semiquantitative XRD analysis. The personnel of the Dept. Inorganic Chemistry (Universidad de Granada) helped with FTIR measurements. We thank F. Gordillo S.L. for donating FG quicklime and R. Ševčı ́k for providing the commercial nanolime.

Figure 6. Degree of carbonation of untransformed (●) and partially transformed (into Ca-ethoxide) commercial CaLoSil nanolime (■) following reaction in ethanol for 14 days at 60 °C. The inset shows the t-evolution of CaCO3 until full carbonation of CaLoSil nanolime partially transformed into Ca-ethoxide. Solid lines are a guide to the eye.

■

(1) Ambrosi, M.; Dei, L.; Giorgi, R.; Neto, C.; Baglioni, P. Colloidal particles of Ca(OH)2: properties and applications to restoration of frescoes. Langmuir 2001, 17, 4251−4255. (2) Giorgi, R.; Baglioni, M.; Berti, D.; Baglioni, P. New methodologies for the conservation of cultural heritage: Micellar solutions, microemulsions, and hydroxide nanoparticles. Acc. Chem. Res. 2010, 43, 695−704. (3) López-Arce, P.; Gomez-Villalba, L. S.; Pinho, L.; Fernández-Valle, M. E.; Alvarez de Buergo, M.; Fort, R. Influence of porosity and relative humidity on consolidation of dolostone with calcium hydroxide nanoparticles: Effectiveness assessment with non-destructive techniques. Mater. Charact. 2010, 61, 168−184. (4) D’Armada, P.; Hirst, E. Nano-lime for consolidation of plaster and stone. J. Architect. Conserv. 2012, 18, 63−80. (5) Chelazzi, D.; Poggi, G.; Jaidar, Y.; Toccafondi, N.; Giorgi, R.; Baglioni, P. Hydroxide nanoparticles for cultural heritage: Consolidation and protection of wall paintings and carbonate materials. J. Colloid Interface Sci. 2013, 392, 42−49. (6) Rodriguez-Navarro, C.; Suzuki, A.; Ruiz-Agudo, E. Alcohol dispersions of calcium hydroxide nanoparticles for stone conservation. Langmuir 2013, 29, 11457−11470. (7) Arizzi, A.; Gomez Villalba, L. S.; Lopez Arce, P.; Cultrone, G.; Fort, R. Lime mortar consolidation with nanostructured calcium hydroxide dispersions: the efficacy of different consolidating products for heritage conservation. Eur. J. Mineral. 2015, 27, 311−323. (8) Baglioni, P.; Chelazzi, D.; Giorgi, R. Nanotechnologies in the Conservation of Cultural Heritage; Springer: New York, 2015. (9) Natali, I.; Tempesti, P.; Carretti, E.; Potenza, M.; Sansoni, S.; Baglioni, P.; Dei, L. Aragonite crystals grown on bones by reaction of CO2 with nanostructured Ca(OH)2 in the presence of collagen. Implications in archaeology and paleontology. Langmuir 2014, 30, 660−668. (10) Poggi, G.; Toccafondi, N.; Melita, L. N.; Knowles, J. C.; Bozec, L.; Giorgi, R.; Baglioni, P. Calcium hydroxide nanoparticles for the conservation of cultural heritage: new formulations for the deacidification of cellulose-based artifacts. Appl. Phys. A: Mater. Sci. Process. 2014, 114, 685−693. (11) Giorgi, R.; Dei, L.; Ceccato, M.; Schettino, C. V.; Baglioni, P. Nanotechnologies for conservation of cultural heritage: paper and canvas deacidification. Langmuir 2002, 18, 8198−8203. (12) Giorgi, R.; Chelazzi, D.; Baglioni, P. Nanoparticles of calcium hydroxide for wood conservation. The deacidification of the Vasa warship. Langmuir 2005, 21, 10743−10748.

thus expected that although the kinetics of the process would be slower, it will however result in a consolidation effect once the partially converted commercial nanolimes are applied for consolidation purposes. This is supported by the reported consolidation efficacy of Ca-alkoxides used for conservation purposes.25,60

■

REFERENCES

CONCLUSIONS

The formation of Ca-alkoxides, which has been generally neglected in previous studies, needs to be considered when applying nanolime treatments for the conservation of cultural heritage. Our results show that in fact Ca(OH)2 particles react with ethanol and isopropanol and partially transform into Caethoxide and Ca-isopropoxide, respectively. The transformation shows Arrhenius behavior and involves the pseudomorphic replacement of Ca(OH)2 particles by Ca-alkoxides. Such a replacement reaction involves a 3D diffusion-controlled deceleratory advancement of the reaction front and progresses via an interface-coupled dissolution−precipitation replacement mechanism. Our results also show that conversion into Caalkoxides significantly reduces the rate of carbonation of Ca(OH)2 particles and induces the formation of metastable vaterite, as opposed to stable calcite which forms in untransformed Ca(OH)2 samples. These effects are observed both in slaked lime putty samples dispersed in alcohol (larger particles and lower surface area) and in a commercial nanolime (smaller particle size and higher surface area). Such effects may be detrimental for the practical application of nanolimes in heritage conservation. Because the transformation of Ca(OH)2 nanoparticles into Ca-alkoxides is T- and t-dependent, freshly prepared alcohol dispersions should be preferred when a fast and effective consolidation is desired. Also, if storage is required prior to application, nanolimes dispersions should be kept at low T because their conversion into Ca-alkoxides displays Arrhenius behavior. 5192

DOI: 10.1021/acs.langmuir.6b01065 Langmuir 2016, 32, 5183−5194

Article

Langmuir

in calcium hydroxide crystallites in concrete. Phys. Status Solidi A 2003, 200, 307−325. (37) Morishita, M.; Kawamura, F.; Kawahara, M.; Yoshimura, M.; Mori, Y.; Sasaki, T. The influence of super saturation on LPE growth of CaN single crystals using the Na flux method. J. Cryst. Growth 2004, 270, 402−408. (38) Putnis, A. Mineral replacement reactions. Rev. Mineral. Geochem. 2009, 70, 87−124. (39) Goel, S. C.; Matchett, M. A.; Chiang, M. Y.; Buhro, W. E. A very large calcium dialkoxide molecular aggregate having a CdI2 core geometry: Ca9(OCH2CH2OMe)18(HOCH2CH2OMe)2. J. Am. Chem. Soc. 1991, 113, 1844−1845. (40) Clark, J. N.; Ihli, J.; Schenk, A. S.; Kim, Y.-Y.; Kulak, A. N.; Campbell, J. M.; Nisbet, G.; Meldrum, F. C.; Robinson, I. K. Threedimensional imaging of dislocation propagation during crystal growth and dissolution. Nat. Mater. 2015, 14, 780−784. (41) Schott, J.; Brantley, S.; Crerar, D.; Guy, C.; Borcsik, M.; Willaime, C. Dissolution kinetics of strained calcite. Geochim. Cosmochim. Acta 1989, 53, 373−382. (42) Barber, D. J.; Wenk, H.-R.; Hirth, G.; Kohlstedt, D. L. Dislocations in minerals. In Dislocations in Solids; Hirth, J. P., Kubin, L., Eds.; Elsevier: Berlin, 2010; pp 171−232. (43) Gómez-Ramírez, R.; Pound, G. M. Nucleation of a second solid phase along dislocations. Metall. Trans. 1973, 4, 1563−1570. (44) Fratini, E.; Page, M. G.; Giorgi, R.; Cölfen, H.; Baglioni, P.; Demé, B.; Zemb, T. Competitive surface adsorption of solvent molecules and compactness of agglomeration in calcium hydroxide nanoparticles. Langmuir 2007, 23, 2330−2338. (45) Kouzu, M.; Kasuno, T.; Tajika, M.; Sugimoto, Y.; Yamanaka, S.; Hidaka, J. Calcium oxide as a solid base catalyst for transesterification of soyben oil and its application to biodiesel production. Fuel 2008, 87, 2798−2806. (46) Sánchez-Cantú, M.; Pérez-Díaz, L. M.; Rosales, R.; Ramírez, E.; Apreza-Sies, A.; Pala-Ramos, I.; Rubio-Rosas, E.; Aguilar-Franco, M.; Valente, J. S. Commercial hydrated lime as a cost-effective solid base for the transesterification of wasted soybean oil with methanol for biodiesel production. Energy Fuels 2011, 25, 3275−3282. (47) Day, R. E.; Parfitt, G. D.; Peacock, J. Vapor adsorption on rutile pretreated with water, ethanol, hexan-1-ol and hexan-1:6-diol. Discuss. Faraday Soc. 1971, 52, 215−225. (48) Di Cosimo, J. I.; Díez, V. K.; Xu, M.; Iglesia, E.; Apesteguia, C. R. Structure and surface catalytic properties of Mg-Al basic oxides. J. Catal. 1998, 178, 499−510. (49) Brantley, S. L. Kinetics of Mineral Dissolution. In Kinetics of Water-Rock Interaction; Brantley, S. L., Kubicki, J. D., White, A. F., Eds.; Springer: New York, 2008; pp 151−210. (50) Carretti, E.; Chelazzi, D.; Rocchigiani, G.; Baglioni, P.; Poggi, G.; Dei, L. Interactions between nanostructured calcium hydroxide and acrylate copolymers: Implications in cultural heritage conservation. Langmuir 2013, 29, 9881−9890. (51) Ruiz-Agudo, E.; Kudłacz, K.; Putnis, C. V.; Putnis, A.; Rodriguez-Navarro, C. Dissolution and carbonation of portlandite [Ca(OH)2] single crystals. Environ. Sci. Technol. 2013, 47, 11342− 11349. (52) Shih, S.-M.; Ho, C.-S.; Song, Y.-S.; Lin, J.-P. Kinetics of the reaction of Ca(OH)2 with CO2 at low temperature. Ind. Eng. Chem. Res. 1999, 38, 1316−1322. (53) Jackson, P.; Parfitt, G. D. Infra-red study of the surface properties of rutile: Adsorption of ethanol, n-butanol and n-hexanol. J. Chem. Soc., Faraday Trans. 1 1972, 68, 1443−1450. (54) Kim, M.; Yan, S.; Salley, S. O.; Ng, K. Y. S. Competitive transestirification of soybean oil with mixed methanol/ethanol over heterogeneous catalysts. Bioresour. Technol. 2010, 101, 4409−4414. (55) Keller, K. S.; Olsson, M. H. M.; Yang, M.; Stipp, S. L. S. Adsorption of ethanol and water on calcite: Dependence on surface geometry and effect on surface behavior. Langmuir 2015, 31, 3847− 3853. (56) Hayashi, T.; Iida, K.; Shinozaki, H.; Sasaki, K. Preparation and properties of Nb-coated barium titanate composite particles by surface

(13) Giorgi, R.; Dei, L.; Baglioni, P. A new method for consolidating wall paintings based on dispersions of lime in alcohol. Stud. Conserv. 2000, 45, 154−161. (14) Salvadori, B.; Dei, L. Synthesis of Ca(OH)2 nanoparticles from diols. Langmuir 2001, 17, 2371−2374. (15) Daniele, V.; Taglieri, G. Synthesis of Ca(OH)2 nanoparticles with the addition of Triton X-100. Protective treatments on natural stones: Preliminary results. J. Cult. Herit. 2012, 13, 40−46. (16) Nanni, A.; Dei, L. Ca(OH)2 Nanoparticles from W/O Microemulsions. Langmuir 2003, 19, 933−938. (17) Ziegenbalg, G. Verfahren zur verfestigenden Behandlung von mineralischen anorganischen Baustoffen. Patent DE10327514B3 (filing date: 17.06.2003; publication date: 20.01.2005). (18) Taglieri, G.; Daniele, V.; Del Re, G.; Volpe, R. A new and original method to produce Ca(OH)2 nanoparticles by using an anion exchange resin. Adv. Nanopart. 2015, 4, 17−24. (19) Gryglewicz, S. Alkaline-earth metal compounds as alcoholysis catalysts for ester oils synthesis. Appl. Catal., A 2000, 192, 23−28. (20) Ruiz-Agudo, E.; Rodriguez-Navarro, C. Microstructure and rheology of lime putty. Langmuir 2010, 26, 3868−3877. (21) Rodriguez-Navarro, C.; Ruiz-Agudo, E.; Ortega-Huertas, M.; Hansen, E. Nanostructure and colloidal behavior of Ca(OH)2: Implications in cultural heritage conservation. Langmuir 2005, 21, 10948−10957. (22) Baglioni, P.; Chelazzi, D.; Giorgi, R.; Carretti, E.; Toccafondi, N.; Jaidar, Y. Commercial Ca(OH)2 nanoparticles for the consolidation of immovable works of art. Appl. Phys. A: Mater. Sci. Process. 2014, 114, 723−732. (23) Perkin, F. M.; Pratt, L. Action of alcohols on metallic calcium. J. Chem. Soc., Trans. 1909, 95, 159−165. (24) Boyle, T. J.; Hernandez-Sanchez, B. A.; Baros, C. M.; Brewer, L. N.; Rodriguez, M. A. Influencing the morphology and phase of calcium ceramic nanoparticles based on the calcium alkoxide precursors’ characteristics. Chem. Mater. 2007, 19, 2016−2026. (25) Favaro, M.; Tomasin, P.; Ossola, F.; Vigato, P. A. A novel approach to consolidation of historical limestone: the calcium alkoxides. Appl. Organomet. Chem. 2008, 22, 698−704. (26) Yuan, Y.; Lee, T. R. Contact angle and wetting phenomena. In Springer Series in Surface Science; Bracco, G., Holst, B., Eds.; Springer: Berlin, 2013; Vol. 51, pp 3−34. (27) Galwey, A. K.; Brown, M. E. Thermal Decomposition of Ionic Solids; Elsevier: Amsterdam, 1999. (28) Vyazovkin, S.; Wight, C. A. Isothermal and nonisothermal reaction kinetics in solids: In search of ways toward consensus. J. Phys. Chem. A 1997, 101, 8279−8284. (29) Beruto, D. T.; Botter, R. Liquid-like H2O adsorption layers to catalyze the Ca(OH)2/CO2 solid−gas reaction and to form a nonprotective solid product layer at 20°C. J. Eur. Ceram. Soc. 2000, 20, 497−503. (30) Liu, X.; Piao, X.; Wang, Y.; Zhu, S. Calcium ethoxide as a solid base catalyst for the transesterification of soybean oil to biodiesel. Energy Fuels 2008, 22, 1313−1317. (31) Dongare, M. K.; Sinha, A. P. B. Thermal analysis of some metal alkoxides. Thermochim. Acta 1982, 57, 37−45. (32) Koga, N.; Yamane, Y. Thermal behaviors of amorphous calcium carbonates prepared in aqueous and ethanol media. J. Therm. Anal. Calorim. 2008, 94, 379−387. (33) Turova, N. Y.; Turevskaya, E. P.; Yanovskaya, M. I.; Yanovsky, A. I.; Kessler, V. G.; Tcheboukov, D. E. Physicochemical approach to the studies of metal alkoxides. Polyhedron 1998, 17, 899−915. (34) Haryanto, A.; Fernando, S.; Murali, N.; Adhikari, S. Current status of hydrogen production techniques by steam reforming of ethanol: A review. Energy Fuels 2005, 19, 2098−2106. (35) Turova, N. Y.; Turevskaya, E. P.; Kessler, V. G.; Yanovsky, A. I.; Struchkov, Y. T. Synthesis, crystal and molecular structure of calcium oxo ethoxide, [Ca6(μ4-O)2(μ3-OEt)4(OEt)4]14EtOH. J. Chem. Soc., Chem. Commun. 1993, 21−23. (36) Harutyunyan, V. S.; Aboyan, E. S.; Monteiro, P. J. M. X-ray diffraction investigations of deformation and dislocation configuration 5193

DOI: 10.1021/acs.langmuir.6b01065 Langmuir 2016, 32, 5183−5194

Article

Langmuir modification with Nb alkoxide in hydrophobic solvent. J. Sol-Gel Sci. Technol. 1999, 16, 159−164. (57) Cooke, D. J.; Gray, R. J.; Sand, K. K.; Stipp, S. L. S. Interaction of ethanol and water with the {10−14} surface of calcite. Langmuir 2010, 26, 14520−14529. (58) Manoli, F.; Dalas, E. Spontaneous precipitation of calcium carbonate in the presence of ethanol, isopropanol and diethylene glycol. J. Cryst. Growth 2000, 218, 359−364. (59) Sand, K. K.; Rodriguez-Blanco, J. D.; Makovicky, E.; Benning, L. G.; Stipp, S. L. S. Crystallization of CaCO3 in water-alcohol mixtures: Spherulicic growth, polymorph stabilization, and morphology change. Cryst. Growth Des. 2012, 12, 842−853. (60) Natali, I.; Tomasin, P.; Becherini, F.; Bernardi, A.; Ciantelli, C.; Favaro, M.; Favoni, O.; Forrat-Perez, V. J.; Olteanu, I. D.; Romero Sanchez, M. D.; Vivarelli, A.; Bonazza, A. Innovative consolidating products for stone materials: field exposure tests as a valid approach for assessing durability. Heritage Sci. 2015, 3, 6.

5194

DOI: 10.1021/acs.langmuir.6b01065 Langmuir 2016, 32, 5183−5194