Facile Intercalation of Organic Molecules into Hydrotalcites by Liquid

Subscriber access provided by NEW YORK MED COLL

Article

Facile intercalation of organic molecules into hydrotalcites by liquid assisted grinding: yield optimization by a chemometric approach Valentina Toson, Eleonora Conterosito, Luca Palin, Enrico Boccaleri, Marco Milanesio, and Valentina Gianotti Cryst. Growth Des., Just Accepted Manuscript • Publication Date (Web): 01 Oct 2015 Downloaded from http://pubs.acs.org on October 9, 2015

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Crystal Growth & Design is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

Facile intercalation of organic molecules into hydrotalcites by liquid assisted grinding: yield optimization by a chemometric approach Valentina Toson,† Eleonora Conterosito,† Luca Palin,†‡ Enrico Boccaleri, † Marco Milanesio,†* Valentina Gianotti†* †

Università del Piemonte Orientale “A. Avogadro” Dipartimento di Scienze ed Innovazione Tecnologica Viale T. Michel 11, 15121 Alessandria, Italy ‡ Nova Res s.r.l., Via Dolores Bello 3, 28100 Novara, Italy, http://www.novares.org KEYWORDS: Layered Double Hydroxides, Mechanochemistry, Reaction yield optimization, Chemometrics, Design of experiment, Simplex

ABSTRACT: Liquid assisted grinding (LAG) method was employed for the preparation of low cost, stable and efficient functional materials, based on organic molecules intercalated into hydrotalcite (LDH). LAG is here exploited to produce hybrid materials with facile preparation methods, low solvent consumption, short reaction times and high yields, and to allow an easy scale up of the synthesis to industrial production. Six molecules were tested to assess potentialities and limitations of LAG. The experiments showed a significant sensitivity to the molecular nature of the intercalant, resulting in different final yields and also different physical forms of the products (powdery vs. pasty materials). With 2-naphtalenesulfonic acid (2-NSA), where the standard recipe gives a yield of about 50%, experimental procedures were optimized by design of experiment (DoE) and Simplex chemometric techniques to find the optimal intercalation conditions, reaching to 76% of yield. A chemometric-driven strategy with wide applicability in material chemistry for high throughput screening and preparation of intercalated compounds was thus proposed.

INTRODUCTION Hydrotalcites, also known as layered double hydroxides (LDHs), are inorganic natural and synthetic anionic clays. The general chemical formula is [M(II)1−xM(III)x (OH)2]x+(An−)x/n·mH2O, where M(II) and M(III) are bivalent and trivalent cations, respectively, usually in a ratio equal to two, octahedrally coordinated to six hydroxyl groups. The inorganic anions, such as CO32-, Cl-, NO3-, compensate the positive charges of the interlayer space.1,2 In principle, anionic exchange allows the intercalation of every negatively charged chemical species, including organic anions,3,4 that can also be inserted directly during LDH preparation, for instance by hydrothermal methods.5,6 Due to their anion exchange capacity, the LDHs have wide applications in many fields: catalysis,7 preparation of pigments,8 removal of waste agents from water,9,10,11,12,13,14 pharmaceutical15,16, and cosmetic formulations,17,18,19,20 stabilizers,21,22,23 rheology modifiers,24,25,26 and optoelectronic devices. 6,27,28 The aim of the present study was the development of facile preparation methods to obtain low cost, stable and efficient materials, based on organic molecules inserted into hydrotalcite. The intercalation yield was maximized through statistical tools to optimize the preparation procedure. In our previous work,19 liquid assisted grinding (LAG) method for fast and clean preparation of organic-intercalated LDHs nanocomposites was developed. The exchange of the inorganic anion (chloride or nitrate) inside the LDH, by an organic molecule, is possible and easy to do in an almost solvent-free environment. Conversely, other several approaches6,9 17,22,26,28 are much more solvent ant time consum-

ing20,27 also with the use of harsh chemicals, such as sulfuric acid.27 To exploit such advantages, LAG is applied in this work for the screening of organic compounds to be intercalated into hydrotalcite. Six organic compounds, shown in Chart 1, were selected as probe molecules and intercalated with the starting recipe proposed by Conterosito et al.19 These molecules show a variety of functionality, shape, size, isomery or physical chemistry features and allowed testing LAG potentialities and limitations.

Chart 1. Structural formulae of the six tested molecules for screening the LAG method. The intercalation yield was optimized by multivariate approaches from the starting recipe employing the factorial de-

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

signs29 and the Simplex algorithm,30 demonstrating their utility in the design of the synthesis and of its optimization. In fact, both approaches allow to shed light on the presence of synergic and/or antagonist effects among the variables that are impossible to evaluate by the classical One Variable At Time (OVAT) approach. These approaches are widely used in several fields such as chromatographic separation optimization and industrial process development31,32 but to our knowledge, the implementation of statistical tools in Material Science for intercalation yield optimization in LDH field is an unprecedented work. The maximum amount of information and the fastest identification of optimal reaction conditions were obtained through the minimum number of experiments.33 EXPERIMENTAL SECTION Chemicals. All chemicals were purchased from Sigma Aldrich and used without further purification, except for carbonate hydrotalcite PURAL® MG 63HT bought from SASOL. Hydrotalcite was used after the substitution of carbonate with nitrate following the procedure described by Iyi et al.34 having formula [Mg0.68Al0.32(OH)2](NO3)0.32·0.37H2O (LDH_NO3). LDH intercalation starting recipe. LAG method was adopted for the intercalation of anionic molecules. The starting intercalation solution was a mixture of 2/3 NaOH (0.5M) and of 1/3 ethanol.19 The guest molecule was mixed and grinded in zirconia ball mill (Retsch mixer mill MM301) with 0.1g of LDH_NO3 in a ratio 1:1 with the exchange sites in LDH (i.e. the NO3- ions,). Then 1ml of the hydroalcoholic prepared solution was added and the grinding carried out for 5min in zirconia ball mortar. The sample was dried in an oven at 50°C for 30min. The amount of liquid used for the intercalation after the preliminary test was increased from 0.519 to 1ml every 0.1g of LDH used to improve LDH wettability: these preparation conditions will be referred from now as “starting recipe”. The starting recipe was then optimized exploiting factorial design and Simplex algorithm,30 varying the experimental factors, i.e. the amount of NaOH and percent of ethanol in solvent. The procedure and other parameters were standardized and kept constant. In particular, MgAl_NO3LDH and guest molecule in neutral form, except for SDS and SDBS, were grinded in a mortar until the complete homogenization. Before the chemometric optimization of 2-NSA intercalation, a preliminary intercalation was done with Na-2-NSA, to establish if the salt form behaves differently from the acid one in the LAG experiment. The Na-2-NSA gave very small yields, probably because of solubility and/or affinity reasons with LDH surface, therefore the 2-NSA neutral acid form was chosen as starting molecule for all DoE and Simplex experiments. X-ray Diffraction. X-ray powder diffraction (XRPD) patterns were obtained on an ARL XTRA48 diffractometer using Cu Kα radiation (λ =1.54062 Å). All powder diffraction spectra were measured in continuous mode using the following conditions: 2ϑ angular range 2-70° for standard measurements; tube power 45kV and 40mA, step size 0.02° 2ϑ, scan rate 1°/min. XRPD patterns were used to assess qualitatively the success of the preparation and to estimate semi quantitatively the reaction yields, calculated by the peak areas of reactants and products. For the fitting of the patterns, instrumental parameters of the profile were determined at first measuring the NIST silicon standard 640e. To account for the broadening of the peak a Lorentzian convolution was

Page 2 of 9

added with a FWHM that varies according to the relation lor_fwhm = c / Cos(Th). Then only the c parameter was refined to account for the broadening of the peak due to the sample. The background was refined with a Chebyshev polynomial. The other two parameters refined were the peak position and intensity. The reduced number of refined parameters assured stability and reproducibility to the fit. A single peak refinement approach, with constrains on layered peak positions belonging to the same phase, was then carried out using Topas TA.35 Its detailed description is given in the results section. Multivariate optimization methods. Experimental design provides a useful tool for analyzing the effect of experimental factors. The two-levels full factorial design (FFD) allows the study of the effect of the principal factors (i.e. the investigated parameters) and of their interactions on the response. The number of experiments required is 2p, p being the number of investigated factors. For each factor, two maximum and minimum values (usually indicated with “+” and “-“respectively) are chosen, and named “levels”. The 2p experiments correspond to all the possible combinations of the two levels of the considered factors. Factorial design theory has been widely described elsewhere33 Simplex optimization is a stepwise multivariate optimization strategy that explores the domain of the variable through a suitable desirability function.30 The experiments are performed one by one. Each step is self-learning depending on the previous ones, except the starting simplex. If p is the number of the experimental parameters to be optimized, the method starts from a set p+1 initial experiments. This set is the starting geometric shape of the Simplex. The optimization is based on the successive projections of the experiment that gives the worst response with respect to the centroid of the other ones in the opposite direction to that experiment. The classical Simplex method stops searching when no better result can be obtained from the projection of p+1 consecutive experiments. In both approaches the investigated response was the reaction yield, calculated from the area underlying the basal peaks, obtained after peak fit of the XRPD patterns.35 Characterization techniques. The amount of carbon, nitrogen, hydrogen and sulfur was detected by using the EuroVector CHNS analyzer “EA3000”. Reaction tube and GC oven temperatures were 980 °C and 100 °C, respectively. The helium flow was 80 mL min-1. Oxygen (12 ml) was injected at 35 kPa. The run time was 400 s and the retention time of the gases were 33.0 s for N2, 52.0 s for CO2, 170.0 s for H2O and 265.0 s for SO2. Atropine sulphate was the standard for the instrument calibration. 0.5 – 1.5 mg of each sample was put into a tin capsule (3.5 x 5 mm) closed outgassed and analyzed. IR analyses were performed on a Fourier transform infrared (FTIR) Nicolet 5700 spectrometer (Thermo Optics) at a resolution of 4 cm-1 in the spectral range from 4000 to 400 cm-1 and 128 scans. The samples were ground in KBr pellet using a sample/KBr weight ratio of 1:10. Thermogravimetric analysis (TGA) was performed on a Setaram SETSYS Evolution instrument under Oxygen (gas flow 20 mL/min), heating the samples from 50 to 800 °C with a rate of 10 °C/min. Thermograms were corrected by subtraction of the background curve. UV-Visible diffuse reflectance spectra were recorded using a Lambda 900-Perkin Elmer spectrophotometer in the spec-


2

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


tral range from 190 to 800 nm. The samples were dispersed in weight ratio 1:20 with Barium Sulfate to attenuate the absorption. The Kubelka-Munk transformation was then applied to the raw data. Steady-state emission spectra were recorded on a Horiba Jobin_Yvon Model IBH FL-322 Fluorolog 3 spectrometer implemented with a 450-W xenon arc lamp, double-grating excitation and emission monochromators (2.1 nm/mm dispersion; 1200 grooves/mm), and a Hamamatsu Model R928 photomultiplier tube. Emission spectra were corrected by standard correction curve. RESULT AND DISCUSSION The potentialities and limitations of LAG intercalation for the production of organo/inorganic LDH composite was evaluated starting from the knowledge acquired from the previous works.16,18,19 Six organic compounds, shown in Chart 1, were selected and intercalated with the starting recipe. Three are aliphatic sulfonic surfactants, sodium dodecyl sulfate (SDS), sodium dodecylbenzenesulfonate (SDBS) and perfluorobutanesulfonic acid (PFBS). Other three are naphthalensulfonic acid compounds, 2-naphthalensulfonic acid (2-NSA), 6-amino-4-hydroxy-2-naphthalenesulfonic acid (A6H4-2-NSA) and 4-amino-3-hydroxy-1naphthalenesulfonic acid (A4H3-1-NSA) characterized by different functionality, shape, size, and isomery. 2-NSA was selected also for its interesting electron transfer features.27 Screening of LAG efficiency. The feasibility of the intercalation of the chosen molecules by LAG method was firstly investigated, applying the starting recipe described in the experimental section. The estimation of the completeness of the intercalation was carried out by a method based on XRPD analysis. In fact in the XRPD pattern the (00l) basal peak is directly related to the interlayer spacing. Thus, the intercalation of organic anions, larger than the NO3-, leads to the appearance at lower angles of the basal peak (00l) of the hybrid LDH phase. Thus the yield of the intercalation reaction was calculated by the presence of the (00l), (002l) and (003l) reflections of the intercalated compounds at lower angles compared to those of the starting LDH_NO3. As already reported by some of us,36 the main peaks of organic-hydrotalcite site can be superimposed to the (003) peak of inorganic LDH’s, such as carbonate and nitrate hydrotalcite. At first, the area of the main peaks of LDH_Guests and LDH_NO3 where calculated by single peak and by Pawley refinement, by Topas TA.35 The yields (% intercalation) were estimated after area calculations by two different methods, taking or not in account such peak overlapping. Single peak fit was the first option and was performed calculating the area of the unreacted (003) peak of the LDH_NO3 and of the (00l) peak of the intercalated compound LDH_Guest. Then, to take into account the above described overlapping, causing a small underestimation of the yield, the intensities of the first six reflections were refined constraining their positions, since they all belong to the (00l) family. Unfortunately this second, in theory more precise method, gave unstable results because of the low crystallinity of the samples and the lacking of a solved crystal structure of the prepared organic (LDH_Guests), and was discarded. The percent yield was then calculated as follow by single peak refinements: Yield%=(LDH_Guest/(LDH_Guest+LDH_NO3))*100 (eq.1) The intercalation screening of six compounds led to different intercalation yields (Table 1), indicating that LAG treatment with starting recipe cannot be applied to all molecules. Depending on the yield results, three categories can be high-

lighted: high with the yields about 100%, medium with yield higher than 50% and low with percentages less than 50%. The XRPD patterns fitted using Topas TA were reported in Figure S1 of the Supporting Information. Table 1. Yield % of intercalation of the LDH_Guest obtained exchanging the six tested molecules with nitrate through the LAG method. Yield (%) 42% 11% 57% 95% 87% 0%

Guest A6H4-2-NSA A4H3-1-NSA 2-NSA SDS SDBS PFBS

Low Low Medium High High -

In the first case, the simple linear structure of SDS leads to the complete exchange, as indicated by the simple visual inspection of XRPD patterns (Figure 1). It is worth noting that this result is obtained in shorter time and with lower environmental impact with respect to literature methods 9,22,37 Although the pasty form of SDBS surfactant makes its handling complicated, it resulted suitable for LAG procedure. (00l)

intensity[au]

Page 3 of 9

(002l)

(003l) (003) LDH_NO3

c (006) LDH_NO3

(00l)

b a

2

4

6

8

10

12

14

16

18

20

22

24

2θ°

Figure 1. Comparison of XRPD patterns of (a) LDH_NO3, (b) LDH_SDBS and (c) LDH_SDS in 2ϑ angular range 225°. 2-NSA was intercalated with intermediate yields (Table 1), as can be also seen by the XRPD, reported in Figure S2. It is a good candidate for chemometric optimization. The remaining three compounds (PFBS, A4H3-1-NSA and A6H42-NSA) were poorly intercalated, resulting completely inappropriate for LAG exchange (XRPD in Figure S3 and S4). The PFBS was demonstrated completely unsuitable for LAG method. The three surfactants (PFBS, SDS and SDBS) showed different behaviors and yields during LAG intercalation, as expected looking to their different features (Chart 1). Furthermore, the two isomers (A4H3-1-NSA and A6H4-1NSA), despite chemically very similar, show different intercalation yields (see the comparison of XRPD in Figure S4 of the ESI). Given the different yields (Table 1) of the six compounds and the pasty form of SDBS, only the intercalation of 2-NSA is worth to be optimized by the application of statistical tools.


3

Crystal Growth & Design Intercalation yield optimization. The Full Factorial Design (FFD) approach was firstly employed to explore the domain of 2-NSA intercalation conditions. The goal is finding the information about the effect of the variables and the features of the experimental domain. The studied variables were the number of moles of NaOH and the amount of solvent (keeping the volume of the final solution constant) at the levels detailed in Table 2. Four experiments were performed at the extreme values (“+” and “-“ as detailed in the experimental section) of the studied variables. In addition three replicates of the experiments were performed in the central values (experiments “0” in Table 2) of the variation range of the variables to obtain the evaluation of the standard error. Table 2. Percent yields obtained in the experiments carried out by the FFD (from 1 to 5) and in the center of variable domain. NaOH (mole) 1.74E-04 (-) 4.95E-04 (+) 1.74E-04 (-) 4.95E-04 (+) 3.35E-04 3.35E-04 3.35E-04

Exp. 1 2 3 4 0 0 0

EtOH (%) 16 (-) 16 (-) 50 (+) 50 (+) 33 33 33

Yield (%) 53 66 27 58 57 57 56

(006)

(00l)

LDH_NO3 +

(002l) + (003)

LDH_NO3

#

* #

#

*

#

*

*

4

*

3

2

1

0 4

6

8

10

12

14

16

18

20

22

obtained yields range from 27 to 66%, suggesting the presence of an interaction effect between the NaOH and EtOH. In particular, EtOH concentration is high, an high amount of NaOH leads to high yields, whilst low NaOH amounts causes low yields. Conversely, when EtOH is employed at low concentrations, high yields are obtained independently on the amount of NaOH used. Simplex method, hence, was used to explore the experimental domain of the high yields region a in a more focused and dynamic way in the direction indicated by the FFD in Table 2. The optimization by Simplex was performed exploring the same variables of the FFD (i.e. NaOH moles and the percent of Ethanol). The procedure and other parameters were standardized, in the same conditions of FFD. Since the number of variable is two, the resulting geometric shape of the initial Simplex for two variables is a triangle, defined in the Cartesian plane with the percent of ethanol on one axis and the moles of NaOH on the other. The 3 vertices of the triangle correspond to the conditions of the chosen starting experiments (experiment 1-3 in Table 3). The first point was the center of the FFD, instead the second and third ones were calculated according to the standard formula30 with a step of 0.3. Table 3. Experiments of the Simplex optimization. Variable values, Yield % and the note about the experiments reflected of simplex, are reported in columns.

The XRPD patterns of the LDH_2-NSA obtained by LAG from a LDH_NO3, in the different FFD experiments are reported in Figure 2.

intensity [a.u.]

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 9

24

2θ [deg.]

Figure 2. XRPD patterns, shown in 2ϑ angular range 3-25°, of LDH_2-NSA samples prepared in FFD condition (exp 14) and in the center of the variable domain (exp 0). The reflections of LDH_2-NSA are indicated with “*” and (00l), (002l), while with “#” the peaks of un-exchange 2-NSA. All the patterns show the characteristic d00l diffraction peak belonging to LDH_2-NSA at 5.04° in 2ϑ, corresponding to 17.5Ȧ, according to Aloisi et al.27 The intercalation is not complete since a residue of the nitrate form is still present as indicated by the peaks centered at 10° and 20° ((00l)LDH_NO3) and also by the presence of the 2-NSA peak (indicated by the ”#”) less intense but still present. The conversion yields estimated by fitting the XRPD pattern, as previously described, are reported in Table 2. The

Exp. 1 2 3 4 5 6 7 8 9

NaOH (mole) 3.35E-04 3.81E-04 3.47E-04 3.93E-04 4.27E-04 4.15E-04 4.61E-04 4.73E-04 5.07E-04

EtOH (%) 33 34 38 39 36 31 32 37 33

Yield (%) 57 68 66 67 72 71 76 73 74

Note Initial simplex Initial simplex Initial simplex Reflec.1 on 2-3 Reflec.3 on 2-4 Reflec.4 on 2-5 Reflec.2 on 5-6 Reflec.6 on 5-7 Reflec.5 on 7-8

The experimental response is again the reaction yield estimated by fitting the XRD patterns reported in Figure S5 of the Supporting Information. The three initial experiments allowed obtaining yields comprised between 57 and 68%. Then, to go further in the optimization process the reflection of the worst experiment (exp.1) was performed according to the procedure reported in Fischer et al.29 The experiment corresponding to the obtained new set of conditions was performed (exp.4). The yield was 67% that represents a better response than the current worst value (exp.1) that was discarded. The next experimental conditions (exp.5) were then obtained by the reflection of the new worst experiment (exp.3) with respect to exp. 2 and 4. The procedure was then carried out until better yield was obtained, as elucidated in Table 3 (exp.6-9). Since the two last experiments did not give any improvement, the optimal conditions giving the maximum yield were obtained. The best experimental conditions resulted those of exp.7, which allowed to intercalate the 2-NSA with the satisfying yield of 76%, larger than the starting recipe (Figure 3). The crystallinity is also improved as can be seen by the peak shape in Figure 3 and estimated by the refined crystal size by Topas TA (45±2 and 27±5 for the optimized and starting recipe respectively). The results confirm the presence of an interaction effect between NaOH and EtOH not detectable. In fact the Simplex


4

Page 5 of 9

method indicates that the yield increases using medium amounts of EtOH and higher amounts of NaOH. (00l)

2-NSA_ simplex7 starting recipe

(006)LDH_NO3 +

intensity(CPS)

*

(002l) + (003)

LDH_NO3

*#

* b

a

4

6

8

10

12

14

16

18

20

22

24

2θ [deg.]

Figure 3. Comparison of XRPD patterns (shown from 2 to 25° 2ϑ angles) of LDH_2-NSA samples obtained in different conditions: a) starting recipe, corresponding to center of FFD; b) experiment 7 of Simplex. The 3D surface plot (reported in Figure 4), built using all the DoE and Simplex experiments, show the direction of climbing of conversion yield. It is worth noting that the surface is not the result of a model but only an extrapolation from all available yields (Tables 2 and 3). Lowering too much the amount of Ethanol limits the swelling of lamellae with a more difficult exchange and consequent low yields. However, a medium amount of ethanol resulted necessary to allow the intercalation, as also observed in our previous works.16,18,19 The amount of NaOH proved to be higher than the one used by Milanesio et al.,18 but it must be noted that an excess of NaOH leads to the formation of a LDH_CO3 phase (see Supporting Information, Figure S6).

ipe also for the intercalation of these two molecules. The comparison of XRPD patterns of samples obtained with starting recipe and that of exp.7 of simplex, were shown in Figures S7 and S8 of the Supporting Information. A6H4-2NSA and A4H3-1-NSA yields increased from 42 and 11% of the starting recipe (Table 2) to 54 e 12%, using the procedure of exp. 7 in Table 3. This demonstrates a limited but in one case interesting “portability” of the optimized recipe for similar compounds. LDH_2-NSA _7simplex characterization. The best sample (exp. 7 in Table 3) was fully characterized by physicalchemical techniques and compared to the material obtained by the exchange in solution by Aloisi et al.27 The characterization of the obtained LDH_2-NSA by TGA, FT-IR and the CHN elemental analysis confirmed (Table 4) that a small nitrate contamination is still present in the LDH_2-NSA. Moreover, since S and N derive from 2-NSA and NO3- respectively, the Sulphur/Nitrogen molar ratio from elemental analysis is directly related to 2-NSA/NO3 ratio, thus allowing to calculate the yield of the reaction. As can be seen in Table 4, a 76/24% ratio is obtained, identical within experimental error to XRPD data from Table 3. This indication validates the method employed to obtain the yields by refining the XRPD data. Table 4. Elemental Analysis results for C, H, N, S % in the LDH_2-NSA_7simplex sample. LDH_2-NSA average Std. dev

N% 1.078 0.065

C% 19.995 0.075

H% 3.719 0.072

S% 7.826 0.084

The FT-IR spectra of LDH_2-NSA and LDH_NO3 were reported in Figure 5. The intense and sharp band at 1384 cm1 and the broad bands at 1763 cm-1 and 825 cm-1 are assigned to vibrational modes of NO3- anions in the LDH interlayer with D3h symmetry. The O-H band, due to the interlayer water and to hydroxyl groups of brucite-like layer, is separated into two components, centered at 3550 and 3450cm-1, corresponding to OH stretching mode. b

Trasmittance[au]

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4. 3D Surface Plot built through an extrapolation of the yield % trend. The variables (NaOH and EtOH) are reported on x-axis and y-axis respectively, while the yields on z-axis; the cold colors indicate the low yields, whilst the hot ones the high yields. The best operative conditions (exp7. in Table 3) were applied to exchange the two naphthalenesulfonic acids, A6H42-NSA and A4H3-1-NSA. The aim was verifying if these conditions, optimized for 2-NSA, could be an improved rec-

a

4000

3750

3500

3250

3000

1750 1500 1250 1000

Wavenumber/cm

750

500

-1

Figure 5. FT-IR spectra of LDH_NO3 (grey line) and LDH_2-NSA_7simplex (black line). The results are shown in two significant range: 4000-2750 cm-1 and 1800-400 cm-1. The absorption bands at 549 and 447cm-1 region correspond to M-O an O-M-O stretching vibrations of the brucite


5

Crystal Growth & Design structure.38,39 In the LDH_2-NSA spectrum, the SO3 symmetric and asymmetric stretching modes at 1180, 1138, 1096 and 1034 cm-1 are attributed to 2-NSA anion intercalated. Several bands in the wavenumber range from 700 to 900 can be assigned to C-H deformations of naphthalene ring.26,40 The stretching bands of nitrate anion still present fall at the same wavenumber of starting LDH_NO3, keeping their D3h symmetry.38 The thermal stability of the compounds was investigated under air flow (Figure 6). The TG profile of LDH_NO3 is characterized by weight losses in two different ranges. The first loss centered at 187°C is ascribed to the chemisorbed and physisorbed water. The second weight loss in the range from 300 to 550°C is associated to different and simultaneous processes, leading to the structure collapse, forming Mg/Al mixed oxide of formula Mg0.68Al0.32O1.16. The main loss of NO3- is rather sharp, and centered at 430°C. Minor NO3- losses occur from 360 to 540°C with multiple steps,36,41,42 superimposed to the de-hydroxylation. The TG trend of the hybrid 2-NSA_LDH shows three distinguishable range of weight loss. Below 180°C a 11.5% loss is associated to surface and interlayer water. At around 400°C there is the overlap of the de-hydroxylation and the depletion of not exchanged nitrate and at 504°C the oxidation of 2-NSA. The 2-NSA loss is rather sharp and centered at 515°C when intercalated into LDH, while to broader losses (at 300 and 490°C) are observed for 2-NSA alone. This confirms that intercalation stabilizes the 2-NSA up to 500°C. 0 16 14

-10

12 -20 10 8

-30

dTG

weight loss%

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

6 -40 4 2

-50

0 -60 100

200

300

400

500

600

700

800

Temperature(°C)

Figure 6. TGA curves collected under oxygen flow in the temperature range from 50 to 800°C of LDH_NO3 (grey curve) and LDH_2-NSA_7simplex (black curve). The differential TG profiles are represented in figure with the same color of TG curves. Given the indications by CHN and TGA analyses, together with initial composition of LDH_NO3, the following formula can be hypothesized for 2-NSA sample: [Mg0.68Al0.32(OH)2](2-NSA)0.24(NO3)0.08·0.37H2O. The optical features of LDH_2-NSA were evaluated through UV-Visible diffuse reflectance and emission spectra recorded and are shown in Figure S9 of the Supporting Information. The DR-UV spectra obtained from LDH_2NSA and 2-NSA are similar, not remarkable change of UV features are showed after intercalation into brucite matrix. Although the rather different almost solvent-free reaction environment, LDH_2-NSA sample is similar to the tradition-

Page 6 of 9

ally synthesized ones27 as demonstrated by XRPD, TGA and DR-UV data. CONCLUSIONS LAG preparation method was used for the facile screening and preparation of several organic-intercalated LDH samples. Six molecules (Chart 1) were tested to assess the feasibility of LAG intercalation. The starting point was the recipe from Conterosito et al.19 In two cases, PFBN and A4H3-1NSA anions, low or no intercalation occurred and the hydrothermal treatment,6 the coprecipitation17 and the exchange in solution for some hours23/days20 seems to be the unique choice for the intercalation into LDH. These molecules proved to be unsuitable for the LAG procedure in the considered operative conditions, probably because of their structure and features. A low yield was obtained with A6H4-2NSA, but it was increased by applying the best conditions of simplex (exp. 7 in Table 3). In the 2-NSA case, where the standard recipe gives a yield of about 57%, the design of experiment approach resulted the best choice to easily and efficiently find the conditions for maximum intercalation yields. SDS was completely intercalated as in the case of coprecipitation9,37 and “memory effect” methods,22 but in a mild and almost solvent-free environment. SDBS, despite the pasty appearance of the sample hinders the use of LAG for the difficulty of manipulating the obtained mixture, leads to interesting yields in the preliminary tests. The exchange of 2-NSA reached a yield of 50% in a very short time and with the low environmental impact of LAG. In literature 2-NSA was intercalated by ethylene glycol (7h at 70°C),28 by titration with sulfuric acid solution,27 by dispersion of salt for some hours/days at RT27 or heating.26 All these procedures are slower and more complicated than LAG method. To improve the yield of this fast, mild and green method, statistical tools were employed to obtain the best conditions for 2-NSA exchange. On one hand, these different yields demonstrated that the experimental conditions of standard recipe are not indistinctly applicable. On the other hand, the multivariate approach allows to improve the yields of the moderately successful intercalation reactions. In the case of 2-NSA, the Factorial Design approach resulted able to improve the yield up to 66%, However, it is not fully suited for yield optimization, because the exploration of the domain is limited and defined a priori on the basis of previous OVAT experiments. Conversely, the simplex algorithm resulted the optimal DoE approach to reach the best conditions in few experiments, also in absence of preliminary experiments, provided that the experimental domain limits can be estimated on the basis of the acidity, and size and shape of the compounds. For LDH_NSA a yield of 76% is reached, starting from 50% of the standard recipe.19 This maximum yield can be considered a physical-chemical limit of the system. Higher yields can be reached by repeating the intercalation putting the so obtained LDH-NSA sample in contact with a fresh NaOH/Ethanol solution.43 The LAG-obtained compounds, despite the rather different almost solvent-free reaction environment resulted similar to the traditionally synthesized ones as demonstrated by XRPD, TGA and DR-UV data. It can be concluded that the search for new intercalated compound can be carried out in two quick steps, to achieve a high throughput approach for the screening and preparation of new materials in large scales, green and mild methods and with yields larger than 70%:


6

Page 7 of 9

Crystal Growth & Design 1)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analysis of LAG efficiency (with standard recipe19) in a series of compounds and classification in low, intermediate or high yield, and definition of the experimental domain (NaOH concentration, Ethanol % in water);

2) Low yields compounds must be discarded, high yields compounds are ready for scale up, intermediate yields compounds can be raised to high yields with only a few (between 5 and 9) LAG experiments, by using the Simplex algorithm. The LAG approach can be generally applied to LDH in particular,43 and in general in all the preparations involving solid-liquid reactions. The chemometric approach for the optimization of the preparation procedures can be easily extended to any reaction in material chemistry.

ASSOCIATED CONTENT Supporting Information. XRPD patterns of LDH_Guest synthetized by starting recipe. XRPD patterns all Simplex experiments. XRPD pattern of LDH_Guest obtained in excess of NaOH. Comparison of XRPD patterns of LDH_A4H3-1-NSA and LDH_A6H4-2NSA performed in different operative condition (starting recipe and exp. 7 of Simplex). Optical characterization of LDH_2-NSA_7simplex.

AUTHOR INFORMATION Corresponding Authors * E-mail: [email protected] (V. Gianotti) for the chemometric-based optimization of the preparation and [email protected] (M. Milanesio) for materials characterization.

Notes The authors declare no competing financial interest.

ACKNOWLEDGMENT Piedmont Region and FINPIEMONTE are acknowledged for POR-FESR 2007/2013 funding, project “DyeHard”. L.P. acknowledges financial contributions by the MIUR project “Multidisciplinary modeling of the structure of layered materials” funded as FIRB in 2012, (code RBFR10CWDA) for funding his bursary.

ABBREVIATIONS LDHs, Layered Double Hydroxides; LAG, Liquid Assisted Grinding; SDS, Sodium Dodecylsulfate; SDBS, Sodium Dodecylbenzenesulfonate; PFBS, Perfluorobutanesulfonic acid; 2NSA, 2-naphthalensulfonic acid; A6H4-2-NSA, 6-amino-4hydroxy-2-naphthalenesulfonic acid; A4H3-1-NSA, 4-amino-3hydroxy-1-naphthalenesulfonic acid; OVAT, One Variable At Time; DoE, Design of Experiments; FFD, Full Factorial Design

REFERENCES

(1) Cavani, F.; Trifirò, F.; Vaccari, A. Catalysis Today 1991, 11, 173. (2) Arizaga, G.G.C.; Satyanarayana, K. G.; Wypych, F. Solid State Ionics 2007, 178, 1143. (3) Ambrogi, V.; Fardella, G.; Grandolini, G.; Perioli, L. Int. J. Pharm. 2001, 220, 23. (4) Ping Xu, Z. P,; Braterman, P. S. J. Phys. Chem. C 2007, 111, 4021. (5) He, Q.; Yin, S.; Sato, T. Phys. Chem. Solids 2004, 65, 395. (6) Kumar, S.; Milanesio, Marchese, L.; Boccaleri, E. Phys. Status Solidi A 2009, 206, 2171.

(7) Figueras, F. Top. Catal. 2004, 29, 189. (8) Costantino, U.; Coletti, N.; Nocchetti, M.; Aloisi, G. G.; Elisei, F. Langmuir 1999, 8, 4454. (9) Bruna, F.; Celis, R.; Real, M.; Cornejo, J. J. Hazardous Mat. 2012, 225, 74. (10) Chuang, Y.H.; Liu, C.H.; Tzou, Y.M.; Chang, J.S.; Chiang, P.N. Colloids Surfaces A Physicochem. Eng. Asp. 2010, 366, 170. (11) Gianotti, V.; Benzi M.; Croce G.; Frascarolo P.; Gosetti F.; Mazzucco E.; Bottaro M.; Gennaro M.C. Chemosphere 2008, 73, 1731. (12) Polati, S.; Angioi S.; Gianotti V.; Gosetti F.; Gennaro M.C. J. Environ. Sci. Health Part B-Pestic. Contam. Agric. Wastes, 2006, 41, 333. (13) Polati, S.; Gosetti F.; Gianotti V.; Gennaro M.C. J. Environ. Sci. Health Part B-Pestic. Contam. Agric. Wastes 2006, 41, 765. (14) Angioi, S.; Polati S.; Roz M.; Rinaudo C.; Gianotti V.; Gennaro M.C. Environ. Pollut. 2005, 134, 35. (15) Costantino, U.; Ambrogi, V.; Nocchetti, M.; Perioli, L. Microporous and Mesoporous Mater 2008, 107, 149. (16) Conterosito, E.; Croce, G.; Palin, L.; Pagano, C.; Perioli, L.; Viterbo, D.; Boccaleri, E.; Paul, G.; Milanesio, M. Phys.Chem. Chem. Phys. 2013, 15, 13418. (17)Khan, S. B.; Liu, C.; Jang, E. S.; Akhtar, K.; Han, H. Mater. Lett. 2011, 65, 2923. (18) Milanesio, M.; Conterosito, E.; Viterbo, D.; Perioli, L.; Croce, G. Cryst. Growth Des. 2010, 10, 4710. (19) Conterosito, E.; Van Beek, W.; Palin, L.; Croce, G.; Perioli, L.; Viterbo, D.; Gatti, G.; Milanesio, M. Cryst. Growth Des. 2013, 13, 1162. (20) Perioli, L.; Nocchetti, M.; Ambrogi, V.; Latterini, L.; Rossi, C.; Costantino, U. Microporous Mesoporous Mater. 2008, 107, 180. (21) Boccaleri, E.; Carniato, F.; Croce, G.; Viterbo, D.; Van Beek, W.; Emerich, H.; Milanesio, M. J. Appl. Cryst. 2007, 40, 684 (22) Costa, F.R.; Leuteritz, A.; Wagenknecht, U.; Auf der Landwehr, U.; Haeussler, L.; Heinrich, G. Appl. Clay Sci. 2009, 7, 14. (23) Zhang, L.; Lin, Y.; Xu, S.; Li, R.; Zheng, X.; Zhang, F. Appl. Clay Sci. 2010, 48, 641. (24) Yang, X.; Zhang, Q. Polym. Int. 2004, 53, 698. (25) Li, Y.; Hou, W. G.; Shen, S. Colloids Surf. A 2009, 350, 109. (26) Raky, L.; Beaudoin, J.J.; Mitchell, L. Cem. Concr. Compos. 2004, 34, 1717. (27) Aloisi, G. G.; Costantino, U.; Elisei, F.; Latterini, L.; Natali, C.; Nocchetti, M. J. Mater. Chem. 2002, 12, 3316. (28) Lei, L.; Kongchao, Z.; Pengfei, L.; Kai. Z. RSC Adv. 2014, 4, 18086. (29) Carlson & Carlson. Design and Optimization in Organic Synthesis, Elsevier, Amsterdam, 1992. (30) Lundstedt, T.; Seifert, E.; Abramo, L.; Thelin, B.; Nyström, A.; Pettersen, J.; Bergman, R.; Chemometr. Intell. Lab. 1998, 42, 3. (31) Marengo, E.; Gennaro, M.C.; Gianotti, V. J. Chromatogr. Sci., 2001, 39, 339. (32) Marengo, E.; Gennaro, M.C.; Gianotti, Chemometr. Intell. Lab. Syst. 2000, 53, 57. (33) Box, G. E. P.; Hunter, W. G.; Hunter, J. S. Statistic for experimenters, Wiley, New York, 1978. (34) Iyi, N.; Yamada, H.; Sasaki, T. Appl. Clay Sci. 2011, 54, 132. (35) Coelho, A. A. J. Appl. Cryst. 2005, 38, 455. (36) Conterosito, E.; Palin, L.; Antonioli, D.; Viterbo, D.; Mugnaioli, E.; Kolb, U.; Perioli, L.; Milanesio, M.; Gianotti, V.; accepted by Chemistry an European Journal, in press. (37) Du, Baoxian. D.; Zhenghong, G.; Zhengping, F. Polym. Degrad. Stab. 2009, 94, 1979. (38) Xu, Z. P.; Zeng, H. C. J. Phys. Chem. B. 2001, 105, 1743. (39) Ahmed, A. A. A.; Talib, Z. A.; Hussein, M. z. B.; Zakaria, A. J. Solid State Chem. 2012, 191, 271.


7


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(40) Socrates, G. Infrared Characteristic Group Frequencies, Wiley. (41) MacKenzie, K. J. D.; Meinhold, R. H.; Sherriffb, B. L.; Xu, Z. J. Mater. Chem. 1993, 3, 1263. (42) Xu, Z. P.; Zeng, H. C. Chem. Mater. 2001, 13, 4564 (43) Conterosito, E.; Benesperi, I.; Toson, V.; Saccone, D.; Barbero, N.; Palin, L.; Gianotti, V.; Barolo, C.; Milanesio, M. High

Page 8 of 9

throughput preparation of additives for photoactive polymers with photovoltaic applications. in preparation.


8

Page 9 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


FOR TABLE OF CONTENTS USE ONLY

Facile intercalation of organic molecules into hydrotalcites by liquid assisted grinding: yield optimization by a chemometric approach Valentina Toson,† Eleonora Conterosito,† Luca Palin,†‡ Enrico Boccaleri, † Marco Milanesio,†* Valentina Gianotti†*

SYNOPSIS Liquid assisted grinding method was employed for the facile and low cost preparation of stable and efficient functional materials, based on organic molecules intercalated into hydrotalcite. A chemometric-driven strategy for reaction yield optimization, exploiting design of experiment (DoE) and simplex approaches, was proposed. The practicality of the method allows an easy scale up of the synthesis to industrial production.


9

Facile Intercalation of Organic Molecules into Hydrotalcites by Liquid

Recommend Documents