- Email: [email protected]
Synthesis of transition metal doped lamellar double hydroxides as base catalysts for acetone aldol condensation
Applied Clay Science 118 (2015) 188–194
Contents lists available at ScienceDirect
Applied Clay Science journal homepage: www.elsevier.com/locate/clay
Research paper
Synthesis of transition metal doped lamellar double hydroxides as base catalysts for acetone aldol condensation M.E. Manríquez a, J.G. Hernández-Cortez b,⁎, J.A. Wang a,⁎, L.F. Chen a, A. Zuñiga-Moreno a, R. Gómez c a b c
ESIQIE, Instituto Politécnico Nacional, UPALM, Col. Zacatenco, 07738 México, DF, México GDMyPQ, Instituto Mexicano del Petróleo, Eje Lázaro Cárdenas 152, 07730 México, DF, México Departamento de Química, Universidad Autónoma Metropolitana-Iztapalapa, San Rafael Atlixco 186, México 09340, DF, Mexico
a r t i c l e
i n f o
Article history: Received 15 May 2015 Received in revised form 14 September 2015 Accepted 2 October 2015 Available online xxxx Keywords: Acetone aldol condensation Lamellar double hydroxides, basicity
a b s t r a c t Transition metals TM (TM = Fe+2, Zn+2, Ni+2 and Cu+2) doped lamellar double hydroxides (LDHs) were prepared by coprecipitation method and characterized with N2 adsorption–desorption isotherms, X-ray diffraction (XRD), Raman spectroscopy, scanning electron microscopy (SEM), and temperature-programmed desorption of CO2 (TPD–CO2). XRD patterns of the dried samples showed that these solids contained the crystallized lamellar double hydroxides (LDHs) after calcination. TPD–CO2 experiments revealed that these LDHs contain both moderate and strong basic sites varying with transition metal dopants. It was found that the gas-phase acetone aldol condensation was a surface basicity and texture-sensitive reaction. The basicity of the catalysts followed the sequence MgZnAlO N MgFeAlO N MgCuAlO N MgNiAlO, which agreed well with the variation of the catalytic activity in the acetone aldol condensation. The selectivity to isophorone and phorone could be correlated to the textural properties of the catalysts. Catalyst with large surface area and bigger pore diameter favored the formation of isophorone. © 2015 Published by Elsevier B.V.
1. Introduction Lamellar double hydroxides (LDHs) attracted great interest because of their wide applications, including their use as catalysts, anion adsorbents and exchangers, medical antacids, and ionic conductors (Dupin et al., 2004; Tanasoi et al., 2009). These materials are composed of a lamellar base structure of magnesium and aluminum hydroxides (Frost et al., 2003; Othman et al., 2006). The general formula of these materials may be represented by [M2+1−xM3+x(OH)2]x+[An−]x / n·yH2O where M2+ are divalent anions (Mg2+, Zn2+, Mn2+, Ni2+, Co2+, Fe2+, etc.), M3+ are trivalent metal ions (Al3+, Cr3+, Fe3+, Co3+, Ga3+, etc.), and An– is the interlayer anion with charge n (Vaccari, 1998). The molecular arrangement and interchange between M2+ ions and M3+ ions in the layered structure generates excess of positive charge which is compensated by the intercalation of anions, generally carbonates and water. The LDHs have structures similar to brucite (Mg(OH)2) sheet-like crystalline structure due to the presence of divalent small cations in close proximity to the highly polarized OH− ions. Each Mg2+ ion has an octahedral coordinated to six hydroxyl ions and various octahedral species share their edges to form an infinite layer. These layers are stacked on one another and seized together by fragile interactions of hydrogen bonds. If any Mg2+ ions are replaced by cations with greater charge but similar ⁎ Corresponding authors. E-mail addresses: [email protected] (J.G. Hernández-Cortez), [email protected] (J.A. Wang).
http://dx.doi.org/10.1016/j.clay.2015.10.002 0169-1317/© 2015 Published by Elsevier B.V.
ionic radius, the layers are positively charged and its electrical neutrality is maintained by anions in the interlayer region together with water molecules (Braterman et al., 1994; Kloprogge et al., 2001; Kloprogge et al., 2005). Therefore, the space between the stacked brucite-like cat− 3− , ion layers is filled with charge compensating anions (CO2− 3 , Cl , NO − SO2− 4 , OH and many others) and water molecules. The synthesis methods, the characterizations of crystalline structures, and the catalytic applications of LDHs-related catalysts were reviewed (Vaccari, 1998). Although the LDH materials can be synthesized by various routes, the most frequently used was the economical method of co-precipitation. The conditions of synthesis (temperature, pH, and metal composition) were reported (Constantino and Pinnavaia, 1994; Climent et al., 2004; Didier et al., 2006; Ajat et al., 2008). Thermal treatments of LDHs led to various modifications in the physicochemical properties of the materials. The most important effect of calcination, at temperatures above 600 °C, was the evaporation of water molecules and carbon dioxide, resulting in the loss of the laminar structure with the consequent formation of aluminum and magnesium mixed oxides. However, it is possible, in some cases, to recover the original LDH structure through rehydration, a phenomenon known as memory effect (Prinetto et al., 2000). Calcination of LDH led to the formation of mixed oxides with stronger basic sites, larger specific area, and increased resistance to sintering. As catalysts, the LDH materials present considerable advantages in low production costs since their synthesis is relatively inexpensive and in terms of environmental pollution, thus they can replace the use
M.E. Manríquez et al. / Applied Clay Science 118 (2015) 188–194
of mineral bases in catalytic reactions. A large number of organic reactions were carried out using LDHs as solid base catalysts and adsorbents (Yamamoto et al., 1995; Schulze et al., 2001; Lima et al., 2014). Several transition metal cations introduced into the LDH can generate Lewis centers, giving rise to catalytic activity. LDHs containing copper and cobalt were reported as successful traps for NOx and SOx, and those containing nickel were active catalysts for the partial oxidation of paraffins (Li et al., 2004). It was reported that BiOCl dispersed on NiFe-LDH was very active for photo-degradation of Rhodamine B dye (Ma et al., 2015). In recent years, the fine chemical industry has developed growing interest in the utilization of solid catalysts in reactions catalyzed by solid bases. The replacement of homogeneous alkaline bases by a solid base catalyst, offers, among others, the advantages of solid reuse and waste stream reduction. The LDHs with basic sites were capable of catalyzing aldol condensation reactions with high conversion (CelayaSanfiz et al., 2015). The aldol condensation of acetone (Ac) produces 4-methyl-4-hydroxy-2-pentanone, which was also known as diacetone alcohol (DAA). Dehydration of DAA leads to 4-methyl-3-pentenone, commonly known as mesityl oxide (MO). Heavier products such as phorones (P) result from additional condensation reactions between MO and acetone. At this point, due to their basic properties, LDHs are promising catalysts for commercial applications in the production of fine chemicals. With this in mind, the use of LDHs for aldol condensation reactions, the synthesis of LDHs solids MgMAlO (M = Zn, Fe, Cu Ni) by the co-precipitation method was reported in the present paper. Their crystalline structures were characterized by XRD, nitrogen adsorption isotherms, SEM, Raman spectroscopy as well as by TPD–CO2 analysis. Aldol condensation of acetone was used as test reaction and the catalytic efficiency and selectivity in function of the surface basicity and textural properties of the transition metal cations incorporated LDHs was explored. 2. Experimental
after outgassing at 400 °C under a pressure of 10−5 Torr for 4 h. X-ray diffraction (XRD) patterns of the solids were obtained by an Empyrean Multi-Purpose Research X-ray diffractometer (PANalytical) with a Cu Kα radiation (λ = 0.15418 nm) source. Scanning electron microscopy (SEM) analysis was made with a Quanta 3D FEG Environmental Scanning Electron Microscope and Focused Ion Beam. For Raman spectroscopy studies, a MicroRaman, Jobin Yvon-Horiba Labram 800 was used. The basicity of the samples was determined by means of temperature programmed desorption of CO2 (TPD–CO2) employing a Micromeritics 2900 device. A CO2–Ar mixture (Praxair Inc.) was employed for the TPD study. The typical procedure consists of catalyst (0.1 g) pretreatment under argon gas flow (30 mL/min) at 200 °C for 2 h, and subsequently, a step of cooling down the sample to room temperature. Finally the sample was saturated under the CO2–Ar mixture. Desorption of carbon dioxide was performed after flushing the system with the carrier gas at an increasing temperature from 200 °C to 500 °C with a heating rate of 10 °C/min. 2.3. Catalytic evaluation Acetone aldol condensation was performed using a fixed-bed reactor at atmospheric pressure. Catalyst sample (200 mg) was loaded in the middle of the reactor and heated in situ at 350 °C in flowing He for 1 h before cooling to the reaction temperature of 300 °C. Acetone was delivered to the reactor at a rate of 0.7 mL h−1. The reaction products were collected in a cold trap and the liquid samples were collected and analyzed with a Varian-3600 CX Gas Chromatograph (Flame Ionization Detector and PONA capillary column). All products were identified by in comparison with their retention times of commercially available pure compounds. The main products detected in the reaction were mesityl oxide (MO), isophorone (IP) and phorone (P). Some deactivation was observed as reaction time increased and the values for activity and selectivity were calculated in the first 60 min. The acetone conversion C and product selectivity Si were calculated as follows:
2.1. Sample preparation The MgAl based hydrotalcite and the transition metal doped lamellar double hydroxides were prepared by co-precipitation method at 65 °C. All samples were prepared as follows: An aqueous solution containing the nitrates of the metallic salts (solution A) was prepared. Another aqueous solution containing KOH and K2CO3 (solution B) was prepared separately. The M2 +/M3 + molar ratio was established to a value of 3 in all cases. Solutions A and B were added drop-wise simultaneously into a glass reactor, maintaining the final pH between 9 and 10. Afterwards, the mixture was aged at 65 °C for 18 h under vigorous stirring. Then the precipitate was washed several times with hot deionized water, and dried at 100 °C in air overnight. The obtained solids were calcined at 450 °C in static air. The metallic nitrate salts employed for the preparation of hydrotalcitelike materials were: M2+: Mg(NO3)2·6H2O, Ni(NO3)2·6H2O, Cu(NO3)2· 6H2O, Zn(NO3)2·6H2O, and M3+: Al(NO3)3·9H2O, and Fe(NO3)3·9H2O. The keys used for synthesized hydrotalcite-like materials are MFA (MgFeAlO), MNA (MgNiAlO), MCA (MgCuAlO), and MZA (MgZnAlO). For comparison purpose, MA (MgAlO) was prepared similarly by coprecipitation using Mg(NO3)2·6H2O and Al(NO3)3·9H2O as Mg and Al precursor; while, the sample MOHA was prepared with Mg(OH)2 and Al(NO3)3·9H2O. 2.2. Characterization Specific surface area, pore volume, and pore size distribution were obtained from nitrogen adsorption–desorption isotherms at −196 °C in an automatic Micromeritics ASAP-2100 analyzer. The chemical composition of solids was determined by inductively coupled plasma atomic emission spectrometry (ICP-AES, Perkin-Elmer model Optima 3200 Dual Vision). The adsorption was carried out using calcined solids,
189
Xn
C ðmol %Þ ¼
Yi Xn 100; Cout þ Y i i i
selectivity Si to compound i: Y Si ðmol %Þ ¼ Xni i
Yi
100
where Cout is acetone mole percent in the outlet of reactor and Yi are yields to the different products. 3. Results and discussion In catalyst preparation, a M2+/M3+ molar ratio of 3 was established in all cases. Elemental chemical analysis data of the LDH samples determined by ICP-AES are summarized in Table 1. Actually, the M2+/M3+ molar ratios in the solids varied from 2.4 to 3.1. The textural properties of the calcined LDHs obtained from the nitrogen adsorption–desorption isotherms were reported in Table 1. All the samples showed surface areas greater than 200 m2/g. It seems that the surface area of the samples was related to the ionic radius of the doping transition metal ions. It was noted that the ionic radius of transition metal M2+ (with 6-fold coordination and low spin state) increased with an order as: Fe2 + (0.61 Å) b Ni2+ (0.69 Å) b Cu2+ (0.73 Å). The surface area of the samples decreased as the ionic radius increased from Fe2 + to Ni2+ and Cu2+. Similarly, the pore volume and pore size gradually decreased as the ionic radius of the divalent cation increased, except for the MZA sample which had a relatively large surface area (234 m2/g), big pore diameter (134 Å), and pore volume (0.78 cm3/g).
190
M.E. Manríquez et al. / Applied Clay Science 118 (2015) 188–194
Table 1 Chemical formula and textural properties of LDHs calcined at 400 °C. Sample MAO MFAO MNAO MCAO MZAO
Metal (wt.%) – 5.5 7.6 7.3 7.6
Chemical formula [Mg0.74Al0.26(OH)2] (CO3)0.13·0.85H20 [Mg0.71Fe0.10Al0.18(OH)2] (CO3)0.14·0.83H20 [Mg0.48Ni0.22Al0.29(OH)2] (CO3)0.15·0.87H20 [Mg0.66Cu0.10Al0.24(OH)2] (CO3)0.12·0.69H20 [Mg0.61Zn0.09Al0.29(OH)2] (CO3)0.14·0.73H20
The XRD patterns of the dried samples were displayed in Fig. 1. The diffraction reflections at around 2θ = 11°, 23°, 35°, 39°, 47°, 61° and 62° corresponded to the crystalline planes of (003), (006), (009), (015), (018), (110), and (113) in these layered double hydroxides (Sumio et al., 2001; Li et al., 2004). The sharp and tense peaks illustrated that the samples contained fine crystalline phases. The XRD patterns of the MOHA sample shown in Fig. 1 indicated a mixture of phases, LDHs and brucite. The brucite phase (Mg(OH)2) was reflected by several XRD sharp peaks marked with *. On the basis of the (003) and (110) peaks, the lattice cell parameters (a = 2d110 and c = 3d003) were calculated and reported in Table 2. The lattice parameters c and a varied with divalent cations in the brucitelike layers, but seemingly each followed an opposite trend. The parameter a, i.e. indicated the shortest distance between two metal cations in the same brucite-like layer. In these samples, the MFN sample was special as it contained both Fe2+ and Fe3+ ions. When the isomorphic substitution of a M2 + by a M3 + cation was made, the cell parameter a increased or decreased, depending mainly on the ionic radii of the cations in octahedral coordination. Compared with the lattice cell parameter a of brucite crystal, the value parameter a of the MNA sample decreased, while for the MFA and MZA, it increased. This change was attributed to an increase of attractive interaction within the brucite-like layer when more Zn2 + and Fe2 + were incorporated (Xu and Zeng, 2001). Thus, a simple way to confirm the isomorphic substitution of divalent and/or trivalent cations is by analyzing the variation of the cell parameter a. These results can be rationalized considering that the ionic radius of Fe3+ is larger than that of Al3+, and in the series of samples with divalent cations M2 +, Zn2 + and Cu2 + sizes are larger than Mg2 +. The differences in c parameters between various LDHs were attributed to the different coulombic attractive forces between the positively charged brucite-like layer and the anion located in the interlayer region. Introducing a cation with higher electronegativity into the brucite structure increased the strength of said attractive forces, and decreased interlayer distances (Sánchez-Cantú et al., 2013).
(
M2+/M3+ molar ratio)
Surface area (m2/g)
Average pore diameter (Å)
VPore (cc/g)
2.9 2.5 2.4 3.1 2.4
253 286 248 204 234
180 127 82 73 134
1.14 0.91 0.51 0.37 0.78
The Raman spectra of the LDHs in the region 200–2000 cm−1 were presented in Fig. 2. Each sample showed a sharp band at approximately 1059 cm−1, which was assigned to vibration of symmetric ν1 CO23 − units in carbonate species bonded to the surface OH, and one with lower intensity at around 1080 cm−1, which was ascribed to the symmetric stretching mode of carbonate anions bonded to interlayer water (Frost and Jagannadha, 2006). In the region between 1200 and 1350 cm−1, vibrations of the carbonate species on the external surface of the LDHs were observed (Kawabata et al., 2005). The band at 1188 cm−1 was attributed to the CO2− 3 antisymmetric stretching vibrations (Frost and Kristy, 2005). The weak band at ~ 1650 cm−1 corresponded to deformation mode of the adsorbed water in the interlayer. The bands observed in the low-frequency region of the spectrum were interpreted as the lattice vibration modes M–O and O–M–O vibrations, in this case, M = Fe2+, Ni2+, Cu2+ and Zn2+, respectively. The Raman spectrum of the sample MFA indicated that the LDH solid contained both Fe2 + and Fe3 + ions in its structure. A sharp band at 1506 cm−1 strongly indicated the presence of Fe3+ ions (Sousa et al., 2000), and the band at about 1927 cm− 1 corresponded to species of Fe2 + (Holgado et al., 1996; Kloprogge et al., 2004.), which suggested the co-existence of both Fe2+ and Fe3+ species in the sample. The presence of Fe2+ ions was due to the reduction of Fe3+ to Fe2+ during the intercalations. The partial reduction of Fe3+ to Fe2+ might be associated with the hydrogen bonding between the interlayer water molecules or hydroxyl groups of the LDH sheets and the Fe3+ cations (Sousa et al., 2000). As assigned above, the bands at 1061 cm− 1 and 1262 cm− 1 were assigned to the vibrations of the carbonate species (CO3)2−, respectively (Miranda et al., 2014). To explore the surface basicity of the catalysts, temperatureprogrammed desorption of CO2 technique was employed. The TPD– CO2 profiles of the catalysts are shown in Fig. 3. Two CO2 desorption peaks were observed in the temperature range from 100 to 500 °C, indicating the different types of basic sites with different basicity strength in the catalysts, mainly associated to Mg2 +–O2 − pairs. The sample MA showed two small peaks, one was between 150 and 220 °C, and another between 300 and 500 °C, Therefore, undoped MA sample contains small number of the basic sites. When the samples doped with transition metal Fe2 +, Ni2 +, and Cu2 +, the TPD–CO2 peaks were gradually increased and the temperature corresponding to the peak maximum between 300 and 450 °C clearly shifted to higher temperature region, indicating that both the number of the basic sites and strength of the basicity increased. In addition, introducing transition metal ions principally enhanced the number of the basic sites with strong strength. It is noted that in the MZA sample, the area of the CO2–TPD peak between 320 °C and 420 °C increased significantly. The temperature corresponding to the peak maximum shifted to lower temperature region in Table 2 Lattice parameters and crystallite size of the LDHs at 120 °C.
Fig. 1. XRD patterns of the different LDH samples.
Sample
d(110)
d(003)
a (Å)
c (Å)
L003a (Å)
L110a (Å)
MA MFA MNA MCA MZA
1.5314 1.5229 1.5290 1.5359 1.5371
7.8200 7.8537 7.8225 7.8430 7.7530
3.0628 3.0458 3.0580 3.0718 3.0742
23.4600 23.5601 23.4675 23.5290 23.2590
296 303 148 207 172
252 254 288 198 417
a
Crystal sizes calculated by the Scherrer equation on the (003) and (110) reflections.
M.E. Manríquez et al. / Applied Clay Science 118 (2015) 188–194
Fig. 2. Raman spectra of the different LDH samples.
comparison with that shown in the sample MFA. Therefore, MZA contains the largest number of basic sites among these catalysts, but its basicity strength was moderate. It was found that a certain number of weak Lewis acid sites were present in the samples as confirmed by infrared characterization of pyridine adsorption (not shown here) because of aluminum in the DLH structure. In combination with the results obtained from TPD–CO2 experiments, we concluded that all the DLH solids contained both weak Lewis acid sites and a big number of basic sites with strength varying from moderate to strong. These acid–base natures may influence their catalytic properties. In Fig. 4, SEM micrographs for the LDH samples were presented. In the sample MCA, Fig. 6a, the plate-type morphologies were observed. The plates were stacked one upon the other and were composed of small agglomerates. The morphologies of the sample MZA were shown in Fig. 6b, where needle formations were observed with an approximate size of 100 nm. Fig. 6c showed some needle-like structures with a length of around 200 nm in the sample MNA, some particles were formed probably by breaking of the needle-like structures. Fig. 6d showed particles with a size of approximately 150 nm. Fig. 5 showed the SEM micrograph of MOHA. This sample consisted of many particles
Fig. 3. Profiles of temperature-programmed desorption of CO2 for different samples.
191
with diameter approximately 150–200 nm and rods with length around 500–600 nm. XRD analysis showed that this sample consisted of two phases, brucite and LDH, Fig. 1. The brucite phase usually showed rod shape (Wang et al., 1998). Herein the rod structures in MOHA sample probably corresponded to brucite phase and the particles were related to LDH phase. The catalytic properties of the calcined LDHs were evaluated in the aldol condensation of acetone (Ac). Fig. 6 showed the conversion of acetone as a function of reaction time. It was seen that the acetone conversion decreased as reaction time increased until reaching stability after 200 min. The initial acetone conversion at 300 °C was 28, 27, 16 and 8 mol% for the MZA, MFA, MCA, and MNA catalysts, respectively. The highest conversion was achieved over the MZA catalyst which showed the stronger surface basicity as well as smaller crystalline size. When the surface basicity was used to explain the catalytic activity, we found that in the transition metal doped catalysts, the catalytic activity was propositional to surface concentration of basic sites. This indicated that surface basicity was one of the key parameters determining the catalytic performance in the acetone aldol condensation. The acetone aldol condensation was a complex reaction in which the selectivity was very sensitive to the nature of the basic sites by considering the fact that the rate-limiting step of the reaction mechanism was the abstraction of the proton H+ from the acetone molecule that would be essentially promoted by the strong basic sites on the solid catalysts. Usually, the acetone aldol condensation in basic medium occurred through several steps. Firstly, basic catalysts subtracted the proton (H+) adjacent to the carbonyl. This, in turn, reacted with the carbonyl group of another acetone (Ac) molecule. The formed diacetone alcohol (DAA) was subsequently dehydrated to produce mesityl oxide (MO). According to the reaction mechanism, selectivity was limited by the production of DAA, so the formation of MO was proportional to the amount of DAA produced. In such case, MO production was strongly affected by the presence of water (see Eq. (1)). Ac ⇔ DAA ⇔ MO þ H2 O
ð1Þ
In the course of 240 min of reaction, the catalytic activity of the catalysts gradually declined. Two possible reasons were probably responsible for the catalyst deactivation: water presence in the reaction steam and coke formation on the catalyst surface. When one molecule of isophorone and mesityl oxide was formed, three molecules of water were produced. Water formation not only increased the partial pressure of water, which inhibited the forward reactions, but also poisoned the basic sites of the catalysts. In the other hand, coke was formed during the reaction as indicated by gray-dark color of the catalyst after the reaction. Chain growth of acetone by aldol condensation took place by interaction of carbonyl oxygen in acetone with a weak Lewis acid site and interaction of α-proton with a strong Lewis base site on the catalyst; as a consequence, electron-acceptor and electron-donor centers were created in the adsorbed molecule (Di Cosimo and Apesteguía, 1998). Strong adsorption of intermediates and subsequent condensation yielded higher non-volatile oligomeric compounds on the surface, blocking the basic sites and finally leading to catalyst deactivation (Chikan et al., 1999). Fig. 7 presented the variation of the mesityl oxide (MO) selectivity as a function of reaction time with different catalysts. The highest selectivity was achieved over the MNA sample following the trend of Ni2+ N Cu2+ N Fe2+ N Zn2+. At low temperatures, the aldol formation is stable for each catalyst; while at high temperatures, the formed aldol is rapidly dehydrated to mesityl oxide (MO) (Sádaba et al., 2011). Additional to the formation of MO, other molecules like the phorone (P) and isophorone (IP) were also obtained in the products. A reaction mechanism based on the analysis of products was displayed in Fig. 8. Phorone formation was limited by the amount of acetone in the reaction medium. However, for the production of isophorone, a suitable reaction condition was required under which dehydration took place (Podrebara
192
M.E. Manríquez et al. / Applied Clay Science 118 (2015) 188–194
Fig. 4. SEM micrographs of the samples: a) MCA; b) MZA; c) MNA; d) MFA.
Fig. 5. SEM micrograph of MOHA with needles morphology.
Fig. 6. Conversion vs time for the aldol condensations of acetone using LDH materials as catalysts. a) MNA; b) MOHA; c) MA; d) MCA; e) MFA; and f) MZA.
M.E. Manríquez et al. / Applied Clay Science 118 (2015) 188–194
Fig. 7. Selectivity to MO in the acetone aldol condensation using LDH materials as catalysts. a) MNA; b) MCA; c) MFA; d) MZA; e) MA; f) MOHA.
et al., 1997; Zamora et al., 2005; Zavoianu et al., 2005; Díez et al., 2006; Carriazo et al., 2007). The selectivity to isophorone was much greater than that to phorone. Fig. 9 showed an increased amount of IP catalyzed with the MA and MFA catalysts. Because the formation of isophorone was closely related to pore diameter and pore volume of the catalysts, catalysts with larger pore diameter and bigger pore volume exhibited higher isophorone selectivity as shown in MA sample, which might be related to the fast diffusivity of the products within the larger pores. Therefore, acetone aldol condensation was also sensitive to textural properties of the catalysts.
193
Fig. 9. Selectivity to isophorone and phorona using LDH materials as catalysts.
4. Conclusions Transition metal TM (TM = Ni, Fe, Cu, and Zn) doped LDHs catalysts were prepared by the co-precipitation method. The LDHs showed good crystallinity increasing with the following sequence of doped cations Fe3+, Ni2+, Cu2+ and Zn2+. All the samples showed surface areas greater than 200 m2/g. The catalytic activity of the catalysts in the aldol condensation reaction was largely governed by the surface basicity. The most active catalyst was Zn doped LDHs which contained the largest number strong basic sites. Over the different catalysts, acetone was converted to 4-methyl-4-hydroxy-2-pentanone (DAA) that could be
Fig. 8. Mechanism of acetone aldol condensation reactions. Ac: acetone; DAA: diacetone alcohol; MO: mesityl oxide; P: phorone; IP: isophorone.
194
M.E. Manríquez et al. / Applied Clay Science 118 (2015) 188–194
dehydrated to produce 4-methyl-3-pentenone (MO), the latter was finally converted to isophorone and phorone. The selectivity to isophorone was much greater than that to phorone which was related to pore diameter and pore volume of the catalysts. Therefore, aldol condensation over transition metal doped LDHs catalysts was a surface basicity and texture-sensitive reaction. Acknowledgments M. E. Manríquez thanks the technical support given by the Center of Nanoscience and Nanotechnology in the National Polytechnic Institute (IPN) in Mexico. The financial support by IPN (Projects SIP-20150553 and SIP-20150554) is acknowledged. References Ajat, M.M.M., Yusoff, K., Hussein, M.Z., 2008. Synthesis of glutamate–zinc–aluminiumlayered double hydroxide nanobiocomposites and cell viability study. Curr. Nanosci. 4, 391–396. Braterman, P.S., Tan, C., Zhao, J., 1994. Orientational effects in the infrared spectrum of the double layer material, magnesium aluminum hydroxide ferrocyanide. Mater. Res. Bull. 29, 1217–1221. Carriazo, D., Martın, C., Rives, V., 2007. An FT-IR study of the adsorption of isopropanol on calcined layered double hydroxides containing isopolymolybdate. Catal. Today 126, 153–161. Celaya-Sanfiz, A., Morales-Vega, N., De Marco, M., Iruretagoyena, D., Mokhtar, M., Bawaked, S.M., Basahel, S.N., Al-Thabaiti, S.A., Alyoubi, A.O., Shaffer, M.S.P., 2015. Self-condensation of acetone over Mg–Al layered double hydroxide supported on multi-walled carbon nanotube catalysts. J. Mol. Catal. A Chem. 398, 50–57. Chikan, V., Molnar, A., Balazsik, K., 1999. One-step synthesis of methyl isobutyl ketone from acetone and hydrogen over Cu-on-MgO catalysts. J. Catal. 184, 134–143. Climent, M.J., Corma, A., Iborra, S., Epping, K., Velty, A., 2004. Increasing the basicity and catalytic activity of hydrotalcites by different synthesis procedures. J. Catal. 225, 316–326. Constantino, V.R.L., Pinnavaia, T.J., 1994. Structure reactivity relationships for basic catalysts derived from a Mg2+/A13+/CO− 3 layered double hydroxide. Catal. Lett. 23, 361–367. Di Cosimo, J.I., Apesteguía, C.R., 1998. Study of the atalyst deactivation in the basecatalyzed oligomerization of acetone. J. Mol. Catal. A Chem. 130, 177–185. Didier, T., Gerardin, C., Durand, R., Coq, B., 2006. Layered double hydroxides: precursors for multifunctional catalysts. Top. Catal. 39, 89–96. Díez, V.K., Apesteguía, C.R., Di, C.J.I., 2006. Aldol condensation of citral with acetone on MgO and alkali-promoted MgO catalysts. J. Catal. 240, 235–244. Dupin, J., Martinez, H., Guimon, C., Dumitriu, E., Fechete, I., 2004. Intercalation compounds of Mg–Al layered double hydroxides with dichlophenac: different methods of preparation and physic-chemical characterization. Appl. Clay Sci. 27 (1–2), 95–106. Frost, R.L., Jagannadha, B., 2006. Thermo Raman Spectroscopic study of natural layered double hydroxides Manaseite. Spectrochim. Acta 65, 553–559. Frost, R.L., Kristy, L., 2005. Raman spectroscopic study of the hydrotalcite desautelsite Mg6Mn2CO3(OH)16·4H2O. Spectrochim. Acta A 61, 2697–2710. Frost, R.L., Ding, Z., Martens, W.N., Johnson, T.E., 2003. Thermal activation of copper oxide based upon the copper hydrotalcite CuxZn6 − xAl2(OH)16(CO3)·4H2O. Thermochim. Acta 398, 167–174. Holgado, M.J., Rives, V., Sanromh, M.S., Malet, P., 1996. Hexacyanoferrate-interlayered hydrotalcite. Solid State Ionics 92, 273–283. Kawabata, T., Shinozuka, Y., Ohidhi, Y., Shihido, T., 2005. Nickel containing Mg–Al hydrotalcite-type anionic clay catalyst for the oxidation of alcohols with molecular oxygen. J. Mol. Catal. A Chem. 236, 206–215.
Kloprogge, J.T., Hickey, L., Frost, R.L., 2001. Heating stage Raman and infrared emission spectroscopic study of the dehydroxylation of synthetic Mg-hydrotalcite. Appl. Clay Sci. 18, 37–49. Kloprogge, J.T., Hickey, L., Frost, R.L., 2005. The effect of varying synthesis conditions on zinc chromium hydrotalcite: a spectroscopic study. Mater. Chem. Phys. 89, 99–109. Kloprogge, J.T., Weier, M., Crespo, I., Ulibarri, M.A., Barriga, C., Rives, V., Martens, W.N., Ray, L., 2004. Intercalation of iron hexacyano complexes in Zn,Al hydrotalcite. Part 2. A mid-infrared and Raman spectroscopic study. J. Solid State Chem. 177, 1382–1387. Li, F., Zhang, L., Evans, D.G., Forano, C., Duan, X., 2004. Structure and thermal evolution of Mg–Al layered double hydroxide containing interlayer organic glyphosate anions. Thermochim. Acta 424, 15–23. Lima, E., Pfeiffer, H., Flores, J., 2014. Some consequences of the fluorination of brucite-like layers in layered double hydroxides: adsorption. Appl. Clay Sci. 88–89, 26–32. Ma, J., Ding, J., Yu, L., Li, L., Kong, Y., Komarneni, S., 2015. BiOCl dispersed on NiFe-LDH leads to enhanced photo-degradation of Rhodamine B dye. Appl. Clay Sci. 109–110, 76–82. Miranda, L.D.L., Bellato, C.R., Fontes, M.P.F., De Almeida, M.F., Milagres, J.L., Minim, L.A., 2014. Preparation and evaluation of hydrotalcite-iron oxide magnetic organocomposite intercalated with surfactants for cationic methylene blue dye. J. Chem. Eng. 254, 88–97. Othman, M.R., Rasid, N.M., Fernando, W.J.N., 2006. Mg–Al hydrotalcite coating on zeolites for improved carbon dioxide adsorption. Chem. Eng. Sci. 61, 1555–1560. Podrebara, G.G., Ng, F.T.T., Rempel, G.L., 1997. A kinetic study of the aldol condensation of acetone using an anion exchange resin catalyst. Chem. Eng. Sci. 52, 2991–3002. Prinetto, F., Ghiotti, G., Graffin, P., Tichit, D., 2000. Synthesis and characterization of sol– gel Mg/Al and Ni/Al layered double hydroxides and comparison with coprecipitated samples. Microporous Mesoporous Mater. 39, 229–247. Sádaba, I., Ojeda, M., Mariscal, R., Richards, R., López- Granados, M., 2011. Mg–Zr mixed oxides for aqueous aldol condensation of furfural with acetone: effect of preparation method and activation temperature. Catal. Today 167, 77–83. Sánchez-Cantú, M., Pérez-Díaz, L.M., Tepale-Ochoa, N., González-Coronel, J., RamosCassellis, M.E., Machorro-Aguirre, D., Valente, J.S., 2013. Green synthesis of hydrocalumite-type compounds and their evaluation in the transesterification of castor bean oil and methanol. Fuel 110, 23–31. Schulze, K., Makowski, W., Chyzy, R., Dziembaj, R., Geismar, G., 2001. Nickel doped hydrotalcites as catalyst precursors for the partial oxidation of light paraffins. Appl. Clay Sci. 18, 59–69. Sousa, M.H., Tourinho, F.A., Rubim, J.C., 2000. Use of Raman micro-spectroscopy in the characterization of MIIFe2O4 (M = Fe, Zn) electric double layer ferrofluids. J. Raman Spectrosc. 31, 185–191. Sumio, A., Satoshi, T., Wataru, O., Yoshio, U., Eiichi, N., 2001. Direct intercalation of amino acids layered double hydroxides by coprecipitation. J. Solid State Chem. 162, 52–62. Tanasoi, S., Tanchoux, N., Urda, A., Tichit, D., Sandulescu, I., Fajula, F., Marcu, I., 2009. New Cu-based mixed oxides obtained from LDH precursors, catalysts for methane total oxidation. Appl. Catal. A Gen. 363 (1–2), 135–142. Vaccari, A., 1998. Preparation and catalytic properties of cationic and anionic clays. Catal. Today 41, 53–71. Wang, J.A., Novaro, O., Bokhimi, X., López, T., Gómez, R., Navarrete, J., Llanos, M.E., López Salinas, E., 1998. Characterizations of the thermal decomposition of brucite prepared by sol–gel technique for synthesis of nanocrystalline MgO. Mater. Lett. 35, 317–323. Xu, Z.P., Zeng, H.C., 2001. Abrupt structural transformation in hydrotalcite-like compounds Mg1 − x Alx(OH)2(NO3)x·nH2O as a continuous function of nitrate anions. J. Phys. Chem. B 105, 1743–1749. Yamamoto, T., Kodama, T., Hasegawa, N., Tsuji, M., Tamaura, T., 1995. Synthesis of hydrotalcite with high layer charge for CO2 adsorbent. Energy Convers. Manag. 36, 637–640. Zamora, M., López, T., Gómez, R., Asomoza, M., Melendez, R., 2005. Acetone gas phase condensation on alkaline metals doped TiO2 sol–gel catalysts. Appl. Surf. Sci. 252, 828–832. Zavoianu, R., Birjega, R., Pavel, O.D., Cruceanu, A., Alifanti, M., 2005. Hydrotalcite like compounds with low Mo-loading active catalysts for selective oxidation of cyclohexene with hydrogen peroxide. Appl. Catal. A Gen. 286, 211–220.
Contents lists available at ScienceDirect
Applied Clay Science journal homepage: www.elsevier.com/locate/clay
Research paper
Synthesis of transition metal doped lamellar double hydroxides as base catalysts for acetone aldol condensation M.E. Manríquez a, J.G. Hernández-Cortez b,⁎, J.A. Wang a,⁎, L.F. Chen a, A. Zuñiga-Moreno a, R. Gómez c a b c
ESIQIE, Instituto Politécnico Nacional, UPALM, Col. Zacatenco, 07738 México, DF, México GDMyPQ, Instituto Mexicano del Petróleo, Eje Lázaro Cárdenas 152, 07730 México, DF, México Departamento de Química, Universidad Autónoma Metropolitana-Iztapalapa, San Rafael Atlixco 186, México 09340, DF, Mexico
a r t i c l e
i n f o
Article history: Received 15 May 2015 Received in revised form 14 September 2015 Accepted 2 October 2015 Available online xxxx Keywords: Acetone aldol condensation Lamellar double hydroxides, basicity
a b s t r a c t Transition metals TM (TM = Fe+2, Zn+2, Ni+2 and Cu+2) doped lamellar double hydroxides (LDHs) were prepared by coprecipitation method and characterized with N2 adsorption–desorption isotherms, X-ray diffraction (XRD), Raman spectroscopy, scanning electron microscopy (SEM), and temperature-programmed desorption of CO2 (TPD–CO2). XRD patterns of the dried samples showed that these solids contained the crystallized lamellar double hydroxides (LDHs) after calcination. TPD–CO2 experiments revealed that these LDHs contain both moderate and strong basic sites varying with transition metal dopants. It was found that the gas-phase acetone aldol condensation was a surface basicity and texture-sensitive reaction. The basicity of the catalysts followed the sequence MgZnAlO N MgFeAlO N MgCuAlO N MgNiAlO, which agreed well with the variation of the catalytic activity in the acetone aldol condensation. The selectivity to isophorone and phorone could be correlated to the textural properties of the catalysts. Catalyst with large surface area and bigger pore diameter favored the formation of isophorone. © 2015 Published by Elsevier B.V.
1. Introduction Lamellar double hydroxides (LDHs) attracted great interest because of their wide applications, including their use as catalysts, anion adsorbents and exchangers, medical antacids, and ionic conductors (Dupin et al., 2004; Tanasoi et al., 2009). These materials are composed of a lamellar base structure of magnesium and aluminum hydroxides (Frost et al., 2003; Othman et al., 2006). The general formula of these materials may be represented by [M2+1−xM3+x(OH)2]x+[An−]x / n·yH2O where M2+ are divalent anions (Mg2+, Zn2+, Mn2+, Ni2+, Co2+, Fe2+, etc.), M3+ are trivalent metal ions (Al3+, Cr3+, Fe3+, Co3+, Ga3+, etc.), and An– is the interlayer anion with charge n (Vaccari, 1998). The molecular arrangement and interchange between M2+ ions and M3+ ions in the layered structure generates excess of positive charge which is compensated by the intercalation of anions, generally carbonates and water. The LDHs have structures similar to brucite (Mg(OH)2) sheet-like crystalline structure due to the presence of divalent small cations in close proximity to the highly polarized OH− ions. Each Mg2+ ion has an octahedral coordinated to six hydroxyl ions and various octahedral species share their edges to form an infinite layer. These layers are stacked on one another and seized together by fragile interactions of hydrogen bonds. If any Mg2+ ions are replaced by cations with greater charge but similar ⁎ Corresponding authors. E-mail addresses: [email protected] (J.G. Hernández-Cortez), [email protected] (J.A. Wang).
http://dx.doi.org/10.1016/j.clay.2015.10.002 0169-1317/© 2015 Published by Elsevier B.V.
ionic radius, the layers are positively charged and its electrical neutrality is maintained by anions in the interlayer region together with water molecules (Braterman et al., 1994; Kloprogge et al., 2001; Kloprogge et al., 2005). Therefore, the space between the stacked brucite-like cat− 3− , ion layers is filled with charge compensating anions (CO2− 3 , Cl , NO − SO2− 4 , OH and many others) and water molecules. The synthesis methods, the characterizations of crystalline structures, and the catalytic applications of LDHs-related catalysts were reviewed (Vaccari, 1998). Although the LDH materials can be synthesized by various routes, the most frequently used was the economical method of co-precipitation. The conditions of synthesis (temperature, pH, and metal composition) were reported (Constantino and Pinnavaia, 1994; Climent et al., 2004; Didier et al., 2006; Ajat et al., 2008). Thermal treatments of LDHs led to various modifications in the physicochemical properties of the materials. The most important effect of calcination, at temperatures above 600 °C, was the evaporation of water molecules and carbon dioxide, resulting in the loss of the laminar structure with the consequent formation of aluminum and magnesium mixed oxides. However, it is possible, in some cases, to recover the original LDH structure through rehydration, a phenomenon known as memory effect (Prinetto et al., 2000). Calcination of LDH led to the formation of mixed oxides with stronger basic sites, larger specific area, and increased resistance to sintering. As catalysts, the LDH materials present considerable advantages in low production costs since their synthesis is relatively inexpensive and in terms of environmental pollution, thus they can replace the use
M.E. Manríquez et al. / Applied Clay Science 118 (2015) 188–194
of mineral bases in catalytic reactions. A large number of organic reactions were carried out using LDHs as solid base catalysts and adsorbents (Yamamoto et al., 1995; Schulze et al., 2001; Lima et al., 2014). Several transition metal cations introduced into the LDH can generate Lewis centers, giving rise to catalytic activity. LDHs containing copper and cobalt were reported as successful traps for NOx and SOx, and those containing nickel were active catalysts for the partial oxidation of paraffins (Li et al., 2004). It was reported that BiOCl dispersed on NiFe-LDH was very active for photo-degradation of Rhodamine B dye (Ma et al., 2015). In recent years, the fine chemical industry has developed growing interest in the utilization of solid catalysts in reactions catalyzed by solid bases. The replacement of homogeneous alkaline bases by a solid base catalyst, offers, among others, the advantages of solid reuse and waste stream reduction. The LDHs with basic sites were capable of catalyzing aldol condensation reactions with high conversion (CelayaSanfiz et al., 2015). The aldol condensation of acetone (Ac) produces 4-methyl-4-hydroxy-2-pentanone, which was also known as diacetone alcohol (DAA). Dehydration of DAA leads to 4-methyl-3-pentenone, commonly known as mesityl oxide (MO). Heavier products such as phorones (P) result from additional condensation reactions between MO and acetone. At this point, due to their basic properties, LDHs are promising catalysts for commercial applications in the production of fine chemicals. With this in mind, the use of LDHs for aldol condensation reactions, the synthesis of LDHs solids MgMAlO (M = Zn, Fe, Cu Ni) by the co-precipitation method was reported in the present paper. Their crystalline structures were characterized by XRD, nitrogen adsorption isotherms, SEM, Raman spectroscopy as well as by TPD–CO2 analysis. Aldol condensation of acetone was used as test reaction and the catalytic efficiency and selectivity in function of the surface basicity and textural properties of the transition metal cations incorporated LDHs was explored. 2. Experimental
after outgassing at 400 °C under a pressure of 10−5 Torr for 4 h. X-ray diffraction (XRD) patterns of the solids were obtained by an Empyrean Multi-Purpose Research X-ray diffractometer (PANalytical) with a Cu Kα radiation (λ = 0.15418 nm) source. Scanning electron microscopy (SEM) analysis was made with a Quanta 3D FEG Environmental Scanning Electron Microscope and Focused Ion Beam. For Raman spectroscopy studies, a MicroRaman, Jobin Yvon-Horiba Labram 800 was used. The basicity of the samples was determined by means of temperature programmed desorption of CO2 (TPD–CO2) employing a Micromeritics 2900 device. A CO2–Ar mixture (Praxair Inc.) was employed for the TPD study. The typical procedure consists of catalyst (0.1 g) pretreatment under argon gas flow (30 mL/min) at 200 °C for 2 h, and subsequently, a step of cooling down the sample to room temperature. Finally the sample was saturated under the CO2–Ar mixture. Desorption of carbon dioxide was performed after flushing the system with the carrier gas at an increasing temperature from 200 °C to 500 °C with a heating rate of 10 °C/min. 2.3. Catalytic evaluation Acetone aldol condensation was performed using a fixed-bed reactor at atmospheric pressure. Catalyst sample (200 mg) was loaded in the middle of the reactor and heated in situ at 350 °C in flowing He for 1 h before cooling to the reaction temperature of 300 °C. Acetone was delivered to the reactor at a rate of 0.7 mL h−1. The reaction products were collected in a cold trap and the liquid samples were collected and analyzed with a Varian-3600 CX Gas Chromatograph (Flame Ionization Detector and PONA capillary column). All products were identified by in comparison with their retention times of commercially available pure compounds. The main products detected in the reaction were mesityl oxide (MO), isophorone (IP) and phorone (P). Some deactivation was observed as reaction time increased and the values for activity and selectivity were calculated in the first 60 min. The acetone conversion C and product selectivity Si were calculated as follows:
2.1. Sample preparation The MgAl based hydrotalcite and the transition metal doped lamellar double hydroxides were prepared by co-precipitation method at 65 °C. All samples were prepared as follows: An aqueous solution containing the nitrates of the metallic salts (solution A) was prepared. Another aqueous solution containing KOH and K2CO3 (solution B) was prepared separately. The M2 +/M3 + molar ratio was established to a value of 3 in all cases. Solutions A and B were added drop-wise simultaneously into a glass reactor, maintaining the final pH between 9 and 10. Afterwards, the mixture was aged at 65 °C for 18 h under vigorous stirring. Then the precipitate was washed several times with hot deionized water, and dried at 100 °C in air overnight. The obtained solids were calcined at 450 °C in static air. The metallic nitrate salts employed for the preparation of hydrotalcitelike materials were: M2+: Mg(NO3)2·6H2O, Ni(NO3)2·6H2O, Cu(NO3)2· 6H2O, Zn(NO3)2·6H2O, and M3+: Al(NO3)3·9H2O, and Fe(NO3)3·9H2O. The keys used for synthesized hydrotalcite-like materials are MFA (MgFeAlO), MNA (MgNiAlO), MCA (MgCuAlO), and MZA (MgZnAlO). For comparison purpose, MA (MgAlO) was prepared similarly by coprecipitation using Mg(NO3)2·6H2O and Al(NO3)3·9H2O as Mg and Al precursor; while, the sample MOHA was prepared with Mg(OH)2 and Al(NO3)3·9H2O. 2.2. Characterization Specific surface area, pore volume, and pore size distribution were obtained from nitrogen adsorption–desorption isotherms at −196 °C in an automatic Micromeritics ASAP-2100 analyzer. The chemical composition of solids was determined by inductively coupled plasma atomic emission spectrometry (ICP-AES, Perkin-Elmer model Optima 3200 Dual Vision). The adsorption was carried out using calcined solids,
189
Xn
C ðmol %Þ ¼
Yi Xn 100; Cout þ Y i i i
selectivity Si to compound i: Y Si ðmol %Þ ¼ Xni i
Yi
100
where Cout is acetone mole percent in the outlet of reactor and Yi are yields to the different products. 3. Results and discussion In catalyst preparation, a M2+/M3+ molar ratio of 3 was established in all cases. Elemental chemical analysis data of the LDH samples determined by ICP-AES are summarized in Table 1. Actually, the M2+/M3+ molar ratios in the solids varied from 2.4 to 3.1. The textural properties of the calcined LDHs obtained from the nitrogen adsorption–desorption isotherms were reported in Table 1. All the samples showed surface areas greater than 200 m2/g. It seems that the surface area of the samples was related to the ionic radius of the doping transition metal ions. It was noted that the ionic radius of transition metal M2+ (with 6-fold coordination and low spin state) increased with an order as: Fe2 + (0.61 Å) b Ni2+ (0.69 Å) b Cu2+ (0.73 Å). The surface area of the samples decreased as the ionic radius increased from Fe2 + to Ni2+ and Cu2+. Similarly, the pore volume and pore size gradually decreased as the ionic radius of the divalent cation increased, except for the MZA sample which had a relatively large surface area (234 m2/g), big pore diameter (134 Å), and pore volume (0.78 cm3/g).
190
M.E. Manríquez et al. / Applied Clay Science 118 (2015) 188–194
Table 1 Chemical formula and textural properties of LDHs calcined at 400 °C. Sample MAO MFAO MNAO MCAO MZAO
Metal (wt.%) – 5.5 7.6 7.3 7.6
Chemical formula [Mg0.74Al0.26(OH)2] (CO3)0.13·0.85H20 [Mg0.71Fe0.10Al0.18(OH)2] (CO3)0.14·0.83H20 [Mg0.48Ni0.22Al0.29(OH)2] (CO3)0.15·0.87H20 [Mg0.66Cu0.10Al0.24(OH)2] (CO3)0.12·0.69H20 [Mg0.61Zn0.09Al0.29(OH)2] (CO3)0.14·0.73H20
The XRD patterns of the dried samples were displayed in Fig. 1. The diffraction reflections at around 2θ = 11°, 23°, 35°, 39°, 47°, 61° and 62° corresponded to the crystalline planes of (003), (006), (009), (015), (018), (110), and (113) in these layered double hydroxides (Sumio et al., 2001; Li et al., 2004). The sharp and tense peaks illustrated that the samples contained fine crystalline phases. The XRD patterns of the MOHA sample shown in Fig. 1 indicated a mixture of phases, LDHs and brucite. The brucite phase (Mg(OH)2) was reflected by several XRD sharp peaks marked with *. On the basis of the (003) and (110) peaks, the lattice cell parameters (a = 2d110 and c = 3d003) were calculated and reported in Table 2. The lattice parameters c and a varied with divalent cations in the brucitelike layers, but seemingly each followed an opposite trend. The parameter a, i.e. indicated the shortest distance between two metal cations in the same brucite-like layer. In these samples, the MFN sample was special as it contained both Fe2+ and Fe3+ ions. When the isomorphic substitution of a M2 + by a M3 + cation was made, the cell parameter a increased or decreased, depending mainly on the ionic radii of the cations in octahedral coordination. Compared with the lattice cell parameter a of brucite crystal, the value parameter a of the MNA sample decreased, while for the MFA and MZA, it increased. This change was attributed to an increase of attractive interaction within the brucite-like layer when more Zn2 + and Fe2 + were incorporated (Xu and Zeng, 2001). Thus, a simple way to confirm the isomorphic substitution of divalent and/or trivalent cations is by analyzing the variation of the cell parameter a. These results can be rationalized considering that the ionic radius of Fe3+ is larger than that of Al3+, and in the series of samples with divalent cations M2 +, Zn2 + and Cu2 + sizes are larger than Mg2 +. The differences in c parameters between various LDHs were attributed to the different coulombic attractive forces between the positively charged brucite-like layer and the anion located in the interlayer region. Introducing a cation with higher electronegativity into the brucite structure increased the strength of said attractive forces, and decreased interlayer distances (Sánchez-Cantú et al., 2013).
(
M2+/M3+ molar ratio)
Surface area (m2/g)
Average pore diameter (Å)
VPore (cc/g)
2.9 2.5 2.4 3.1 2.4
253 286 248 204 234
180 127 82 73 134
1.14 0.91 0.51 0.37 0.78
The Raman spectra of the LDHs in the region 200–2000 cm−1 were presented in Fig. 2. Each sample showed a sharp band at approximately 1059 cm−1, which was assigned to vibration of symmetric ν1 CO23 − units in carbonate species bonded to the surface OH, and one with lower intensity at around 1080 cm−1, which was ascribed to the symmetric stretching mode of carbonate anions bonded to interlayer water (Frost and Jagannadha, 2006). In the region between 1200 and 1350 cm−1, vibrations of the carbonate species on the external surface of the LDHs were observed (Kawabata et al., 2005). The band at 1188 cm−1 was attributed to the CO2− 3 antisymmetric stretching vibrations (Frost and Kristy, 2005). The weak band at ~ 1650 cm−1 corresponded to deformation mode of the adsorbed water in the interlayer. The bands observed in the low-frequency region of the spectrum were interpreted as the lattice vibration modes M–O and O–M–O vibrations, in this case, M = Fe2+, Ni2+, Cu2+ and Zn2+, respectively. The Raman spectrum of the sample MFA indicated that the LDH solid contained both Fe2 + and Fe3 + ions in its structure. A sharp band at 1506 cm−1 strongly indicated the presence of Fe3+ ions (Sousa et al., 2000), and the band at about 1927 cm− 1 corresponded to species of Fe2 + (Holgado et al., 1996; Kloprogge et al., 2004.), which suggested the co-existence of both Fe2+ and Fe3+ species in the sample. The presence of Fe2+ ions was due to the reduction of Fe3+ to Fe2+ during the intercalations. The partial reduction of Fe3+ to Fe2+ might be associated with the hydrogen bonding between the interlayer water molecules or hydroxyl groups of the LDH sheets and the Fe3+ cations (Sousa et al., 2000). As assigned above, the bands at 1061 cm− 1 and 1262 cm− 1 were assigned to the vibrations of the carbonate species (CO3)2−, respectively (Miranda et al., 2014). To explore the surface basicity of the catalysts, temperatureprogrammed desorption of CO2 technique was employed. The TPD– CO2 profiles of the catalysts are shown in Fig. 3. Two CO2 desorption peaks were observed in the temperature range from 100 to 500 °C, indicating the different types of basic sites with different basicity strength in the catalysts, mainly associated to Mg2 +–O2 − pairs. The sample MA showed two small peaks, one was between 150 and 220 °C, and another between 300 and 500 °C, Therefore, undoped MA sample contains small number of the basic sites. When the samples doped with transition metal Fe2 +, Ni2 +, and Cu2 +, the TPD–CO2 peaks were gradually increased and the temperature corresponding to the peak maximum between 300 and 450 °C clearly shifted to higher temperature region, indicating that both the number of the basic sites and strength of the basicity increased. In addition, introducing transition metal ions principally enhanced the number of the basic sites with strong strength. It is noted that in the MZA sample, the area of the CO2–TPD peak between 320 °C and 420 °C increased significantly. The temperature corresponding to the peak maximum shifted to lower temperature region in Table 2 Lattice parameters and crystallite size of the LDHs at 120 °C.
Fig. 1. XRD patterns of the different LDH samples.
Sample
d(110)
d(003)
a (Å)
c (Å)
L003a (Å)
L110a (Å)
MA MFA MNA MCA MZA
1.5314 1.5229 1.5290 1.5359 1.5371
7.8200 7.8537 7.8225 7.8430 7.7530
3.0628 3.0458 3.0580 3.0718 3.0742
23.4600 23.5601 23.4675 23.5290 23.2590
296 303 148 207 172
252 254 288 198 417
a
Crystal sizes calculated by the Scherrer equation on the (003) and (110) reflections.
M.E. Manríquez et al. / Applied Clay Science 118 (2015) 188–194
Fig. 2. Raman spectra of the different LDH samples.
comparison with that shown in the sample MFA. Therefore, MZA contains the largest number of basic sites among these catalysts, but its basicity strength was moderate. It was found that a certain number of weak Lewis acid sites were present in the samples as confirmed by infrared characterization of pyridine adsorption (not shown here) because of aluminum in the DLH structure. In combination with the results obtained from TPD–CO2 experiments, we concluded that all the DLH solids contained both weak Lewis acid sites and a big number of basic sites with strength varying from moderate to strong. These acid–base natures may influence their catalytic properties. In Fig. 4, SEM micrographs for the LDH samples were presented. In the sample MCA, Fig. 6a, the plate-type morphologies were observed. The plates were stacked one upon the other and were composed of small agglomerates. The morphologies of the sample MZA were shown in Fig. 6b, where needle formations were observed with an approximate size of 100 nm. Fig. 6c showed some needle-like structures with a length of around 200 nm in the sample MNA, some particles were formed probably by breaking of the needle-like structures. Fig. 6d showed particles with a size of approximately 150 nm. Fig. 5 showed the SEM micrograph of MOHA. This sample consisted of many particles
Fig. 3. Profiles of temperature-programmed desorption of CO2 for different samples.
191
with diameter approximately 150–200 nm and rods with length around 500–600 nm. XRD analysis showed that this sample consisted of two phases, brucite and LDH, Fig. 1. The brucite phase usually showed rod shape (Wang et al., 1998). Herein the rod structures in MOHA sample probably corresponded to brucite phase and the particles were related to LDH phase. The catalytic properties of the calcined LDHs were evaluated in the aldol condensation of acetone (Ac). Fig. 6 showed the conversion of acetone as a function of reaction time. It was seen that the acetone conversion decreased as reaction time increased until reaching stability after 200 min. The initial acetone conversion at 300 °C was 28, 27, 16 and 8 mol% for the MZA, MFA, MCA, and MNA catalysts, respectively. The highest conversion was achieved over the MZA catalyst which showed the stronger surface basicity as well as smaller crystalline size. When the surface basicity was used to explain the catalytic activity, we found that in the transition metal doped catalysts, the catalytic activity was propositional to surface concentration of basic sites. This indicated that surface basicity was one of the key parameters determining the catalytic performance in the acetone aldol condensation. The acetone aldol condensation was a complex reaction in which the selectivity was very sensitive to the nature of the basic sites by considering the fact that the rate-limiting step of the reaction mechanism was the abstraction of the proton H+ from the acetone molecule that would be essentially promoted by the strong basic sites on the solid catalysts. Usually, the acetone aldol condensation in basic medium occurred through several steps. Firstly, basic catalysts subtracted the proton (H+) adjacent to the carbonyl. This, in turn, reacted with the carbonyl group of another acetone (Ac) molecule. The formed diacetone alcohol (DAA) was subsequently dehydrated to produce mesityl oxide (MO). According to the reaction mechanism, selectivity was limited by the production of DAA, so the formation of MO was proportional to the amount of DAA produced. In such case, MO production was strongly affected by the presence of water (see Eq. (1)). Ac ⇔ DAA ⇔ MO þ H2 O
ð1Þ
In the course of 240 min of reaction, the catalytic activity of the catalysts gradually declined. Two possible reasons were probably responsible for the catalyst deactivation: water presence in the reaction steam and coke formation on the catalyst surface. When one molecule of isophorone and mesityl oxide was formed, three molecules of water were produced. Water formation not only increased the partial pressure of water, which inhibited the forward reactions, but also poisoned the basic sites of the catalysts. In the other hand, coke was formed during the reaction as indicated by gray-dark color of the catalyst after the reaction. Chain growth of acetone by aldol condensation took place by interaction of carbonyl oxygen in acetone with a weak Lewis acid site and interaction of α-proton with a strong Lewis base site on the catalyst; as a consequence, electron-acceptor and electron-donor centers were created in the adsorbed molecule (Di Cosimo and Apesteguía, 1998). Strong adsorption of intermediates and subsequent condensation yielded higher non-volatile oligomeric compounds on the surface, blocking the basic sites and finally leading to catalyst deactivation (Chikan et al., 1999). Fig. 7 presented the variation of the mesityl oxide (MO) selectivity as a function of reaction time with different catalysts. The highest selectivity was achieved over the MNA sample following the trend of Ni2+ N Cu2+ N Fe2+ N Zn2+. At low temperatures, the aldol formation is stable for each catalyst; while at high temperatures, the formed aldol is rapidly dehydrated to mesityl oxide (MO) (Sádaba et al., 2011). Additional to the formation of MO, other molecules like the phorone (P) and isophorone (IP) were also obtained in the products. A reaction mechanism based on the analysis of products was displayed in Fig. 8. Phorone formation was limited by the amount of acetone in the reaction medium. However, for the production of isophorone, a suitable reaction condition was required under which dehydration took place (Podrebara
192
M.E. Manríquez et al. / Applied Clay Science 118 (2015) 188–194
Fig. 4. SEM micrographs of the samples: a) MCA; b) MZA; c) MNA; d) MFA.
Fig. 5. SEM micrograph of MOHA with needles morphology.
Fig. 6. Conversion vs time for the aldol condensations of acetone using LDH materials as catalysts. a) MNA; b) MOHA; c) MA; d) MCA; e) MFA; and f) MZA.
M.E. Manríquez et al. / Applied Clay Science 118 (2015) 188–194
Fig. 7. Selectivity to MO in the acetone aldol condensation using LDH materials as catalysts. a) MNA; b) MCA; c) MFA; d) MZA; e) MA; f) MOHA.
et al., 1997; Zamora et al., 2005; Zavoianu et al., 2005; Díez et al., 2006; Carriazo et al., 2007). The selectivity to isophorone was much greater than that to phorone. Fig. 9 showed an increased amount of IP catalyzed with the MA and MFA catalysts. Because the formation of isophorone was closely related to pore diameter and pore volume of the catalysts, catalysts with larger pore diameter and bigger pore volume exhibited higher isophorone selectivity as shown in MA sample, which might be related to the fast diffusivity of the products within the larger pores. Therefore, acetone aldol condensation was also sensitive to textural properties of the catalysts.
193
Fig. 9. Selectivity to isophorone and phorona using LDH materials as catalysts.
4. Conclusions Transition metal TM (TM = Ni, Fe, Cu, and Zn) doped LDHs catalysts were prepared by the co-precipitation method. The LDHs showed good crystallinity increasing with the following sequence of doped cations Fe3+, Ni2+, Cu2+ and Zn2+. All the samples showed surface areas greater than 200 m2/g. The catalytic activity of the catalysts in the aldol condensation reaction was largely governed by the surface basicity. The most active catalyst was Zn doped LDHs which contained the largest number strong basic sites. Over the different catalysts, acetone was converted to 4-methyl-4-hydroxy-2-pentanone (DAA) that could be
Fig. 8. Mechanism of acetone aldol condensation reactions. Ac: acetone; DAA: diacetone alcohol; MO: mesityl oxide; P: phorone; IP: isophorone.
194
M.E. Manríquez et al. / Applied Clay Science 118 (2015) 188–194
dehydrated to produce 4-methyl-3-pentenone (MO), the latter was finally converted to isophorone and phorone. The selectivity to isophorone was much greater than that to phorone which was related to pore diameter and pore volume of the catalysts. Therefore, aldol condensation over transition metal doped LDHs catalysts was a surface basicity and texture-sensitive reaction. Acknowledgments M. E. Manríquez thanks the technical support given by the Center of Nanoscience and Nanotechnology in the National Polytechnic Institute (IPN) in Mexico. The financial support by IPN (Projects SIP-20150553 and SIP-20150554) is acknowledged. References Ajat, M.M.M., Yusoff, K., Hussein, M.Z., 2008. Synthesis of glutamate–zinc–aluminiumlayered double hydroxide nanobiocomposites and cell viability study. Curr. Nanosci. 4, 391–396. Braterman, P.S., Tan, C., Zhao, J., 1994. Orientational effects in the infrared spectrum of the double layer material, magnesium aluminum hydroxide ferrocyanide. Mater. Res. Bull. 29, 1217–1221. Carriazo, D., Martın, C., Rives, V., 2007. An FT-IR study of the adsorption of isopropanol on calcined layered double hydroxides containing isopolymolybdate. Catal. Today 126, 153–161. Celaya-Sanfiz, A., Morales-Vega, N., De Marco, M., Iruretagoyena, D., Mokhtar, M., Bawaked, S.M., Basahel, S.N., Al-Thabaiti, S.A., Alyoubi, A.O., Shaffer, M.S.P., 2015. Self-condensation of acetone over Mg–Al layered double hydroxide supported on multi-walled carbon nanotube catalysts. J. Mol. Catal. A Chem. 398, 50–57. Chikan, V., Molnar, A., Balazsik, K., 1999. One-step synthesis of methyl isobutyl ketone from acetone and hydrogen over Cu-on-MgO catalysts. J. Catal. 184, 134–143. Climent, M.J., Corma, A., Iborra, S., Epping, K., Velty, A., 2004. Increasing the basicity and catalytic activity of hydrotalcites by different synthesis procedures. J. Catal. 225, 316–326. Constantino, V.R.L., Pinnavaia, T.J., 1994. Structure reactivity relationships for basic catalysts derived from a Mg2+/A13+/CO− 3 layered double hydroxide. Catal. Lett. 23, 361–367. Di Cosimo, J.I., Apesteguía, C.R., 1998. Study of the atalyst deactivation in the basecatalyzed oligomerization of acetone. J. Mol. Catal. A Chem. 130, 177–185. Didier, T., Gerardin, C., Durand, R., Coq, B., 2006. Layered double hydroxides: precursors for multifunctional catalysts. Top. Catal. 39, 89–96. Díez, V.K., Apesteguía, C.R., Di, C.J.I., 2006. Aldol condensation of citral with acetone on MgO and alkali-promoted MgO catalysts. J. Catal. 240, 235–244. Dupin, J., Martinez, H., Guimon, C., Dumitriu, E., Fechete, I., 2004. Intercalation compounds of Mg–Al layered double hydroxides with dichlophenac: different methods of preparation and physic-chemical characterization. Appl. Clay Sci. 27 (1–2), 95–106. Frost, R.L., Jagannadha, B., 2006. Thermo Raman Spectroscopic study of natural layered double hydroxides Manaseite. Spectrochim. Acta 65, 553–559. Frost, R.L., Kristy, L., 2005. Raman spectroscopic study of the hydrotalcite desautelsite Mg6Mn2CO3(OH)16·4H2O. Spectrochim. Acta A 61, 2697–2710. Frost, R.L., Ding, Z., Martens, W.N., Johnson, T.E., 2003. Thermal activation of copper oxide based upon the copper hydrotalcite CuxZn6 − xAl2(OH)16(CO3)·4H2O. Thermochim. Acta 398, 167–174. Holgado, M.J., Rives, V., Sanromh, M.S., Malet, P., 1996. Hexacyanoferrate-interlayered hydrotalcite. Solid State Ionics 92, 273–283. Kawabata, T., Shinozuka, Y., Ohidhi, Y., Shihido, T., 2005. Nickel containing Mg–Al hydrotalcite-type anionic clay catalyst for the oxidation of alcohols with molecular oxygen. J. Mol. Catal. A Chem. 236, 206–215.
Kloprogge, J.T., Hickey, L., Frost, R.L., 2001. Heating stage Raman and infrared emission spectroscopic study of the dehydroxylation of synthetic Mg-hydrotalcite. Appl. Clay Sci. 18, 37–49. Kloprogge, J.T., Hickey, L., Frost, R.L., 2005. The effect of varying synthesis conditions on zinc chromium hydrotalcite: a spectroscopic study. Mater. Chem. Phys. 89, 99–109. Kloprogge, J.T., Weier, M., Crespo, I., Ulibarri, M.A., Barriga, C., Rives, V., Martens, W.N., Ray, L., 2004. Intercalation of iron hexacyano complexes in Zn,Al hydrotalcite. Part 2. A mid-infrared and Raman spectroscopic study. J. Solid State Chem. 177, 1382–1387. Li, F., Zhang, L., Evans, D.G., Forano, C., Duan, X., 2004. Structure and thermal evolution of Mg–Al layered double hydroxide containing interlayer organic glyphosate anions. Thermochim. Acta 424, 15–23. Lima, E., Pfeiffer, H., Flores, J., 2014. Some consequences of the fluorination of brucite-like layers in layered double hydroxides: adsorption. Appl. Clay Sci. 88–89, 26–32. Ma, J., Ding, J., Yu, L., Li, L., Kong, Y., Komarneni, S., 2015. BiOCl dispersed on NiFe-LDH leads to enhanced photo-degradation of Rhodamine B dye. Appl. Clay Sci. 109–110, 76–82. Miranda, L.D.L., Bellato, C.R., Fontes, M.P.F., De Almeida, M.F., Milagres, J.L., Minim, L.A., 2014. Preparation and evaluation of hydrotalcite-iron oxide magnetic organocomposite intercalated with surfactants for cationic methylene blue dye. J. Chem. Eng. 254, 88–97. Othman, M.R., Rasid, N.M., Fernando, W.J.N., 2006. Mg–Al hydrotalcite coating on zeolites for improved carbon dioxide adsorption. Chem. Eng. Sci. 61, 1555–1560. Podrebara, G.G., Ng, F.T.T., Rempel, G.L., 1997. A kinetic study of the aldol condensation of acetone using an anion exchange resin catalyst. Chem. Eng. Sci. 52, 2991–3002. Prinetto, F., Ghiotti, G., Graffin, P., Tichit, D., 2000. Synthesis and characterization of sol– gel Mg/Al and Ni/Al layered double hydroxides and comparison with coprecipitated samples. Microporous Mesoporous Mater. 39, 229–247. Sádaba, I., Ojeda, M., Mariscal, R., Richards, R., López- Granados, M., 2011. Mg–Zr mixed oxides for aqueous aldol condensation of furfural with acetone: effect of preparation method and activation temperature. Catal. Today 167, 77–83. Sánchez-Cantú, M., Pérez-Díaz, L.M., Tepale-Ochoa, N., González-Coronel, J., RamosCassellis, M.E., Machorro-Aguirre, D., Valente, J.S., 2013. Green synthesis of hydrocalumite-type compounds and their evaluation in the transesterification of castor bean oil and methanol. Fuel 110, 23–31. Schulze, K., Makowski, W., Chyzy, R., Dziembaj, R., Geismar, G., 2001. Nickel doped hydrotalcites as catalyst precursors for the partial oxidation of light paraffins. Appl. Clay Sci. 18, 59–69. Sousa, M.H., Tourinho, F.A., Rubim, J.C., 2000. Use of Raman micro-spectroscopy in the characterization of MIIFe2O4 (M = Fe, Zn) electric double layer ferrofluids. J. Raman Spectrosc. 31, 185–191. Sumio, A., Satoshi, T., Wataru, O., Yoshio, U., Eiichi, N., 2001. Direct intercalation of amino acids layered double hydroxides by coprecipitation. J. Solid State Chem. 162, 52–62. Tanasoi, S., Tanchoux, N., Urda, A., Tichit, D., Sandulescu, I., Fajula, F., Marcu, I., 2009. New Cu-based mixed oxides obtained from LDH precursors, catalysts for methane total oxidation. Appl. Catal. A Gen. 363 (1–2), 135–142. Vaccari, A., 1998. Preparation and catalytic properties of cationic and anionic clays. Catal. Today 41, 53–71. Wang, J.A., Novaro, O., Bokhimi, X., López, T., Gómez, R., Navarrete, J., Llanos, M.E., López Salinas, E., 1998. Characterizations of the thermal decomposition of brucite prepared by sol–gel technique for synthesis of nanocrystalline MgO. Mater. Lett. 35, 317–323. Xu, Z.P., Zeng, H.C., 2001. Abrupt structural transformation in hydrotalcite-like compounds Mg1 − x Alx(OH)2(NO3)x·nH2O as a continuous function of nitrate anions. J. Phys. Chem. B 105, 1743–1749. Yamamoto, T., Kodama, T., Hasegawa, N., Tsuji, M., Tamaura, T., 1995. Synthesis of hydrotalcite with high layer charge for CO2 adsorbent. Energy Convers. Manag. 36, 637–640. Zamora, M., López, T., Gómez, R., Asomoza, M., Melendez, R., 2005. Acetone gas phase condensation on alkaline metals doped TiO2 sol–gel catalysts. Appl. Surf. Sci. 252, 828–832. Zavoianu, R., Birjega, R., Pavel, O.D., Cruceanu, A., Alifanti, M., 2005. Hydrotalcite like compounds with low Mo-loading active catalysts for selective oxidation of cyclohexene with hydrogen peroxide. Appl. Catal. A Gen. 286, 211–220.