Efficient magnetic recoverable acid-functionalized-carbon catalysts for starch valorization to multiple bio-chemicals

G Model

ARTICLE IN PRESS

CATTOD-10316; No. of Pages 11

Catalysis Today xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Catalysis Today journal homepage: www.elsevier.com/locate/cattod

Efficient magnetic recoverable acid-functionalized-carbon catalysts for starch valorization to multiple bio-chemicals Iunia Podolean a , Florin Anita a , Hermenegildo García b , Vasile I. Parvulescu a,∗ , Simona M. Coman a,∗ a Department of Organic Chemistry, Biochemistry and Catalysis, Faculty of Chemistry, University of Bucharest, Regina Elisabeta Blvd., no. 4-12, Bucharest 030016, Romania b Departamento de Química, Instituto Universitario de Tecnología Química (CSIC-UPV), Avenida de los Naranjos S/N, 46022 Valencia, Spain

a r t i c l e

i n f o

Article history: Received 30 December 2015 Received in revised form 14 June 2016 Accepted 11 July 2016 Available online xxx Keywords: Triflate Sulphonic sites Niobia Carbon-based carrier Magnetite Starch

a b s t r a c t A series of solid acid catalysts were prepared by functionalization of graphene-oxide (GO) and MWCNToxide (MWCNTO) with triflic and sulfonic acids affording a triflate-functionalized graphene oxide (GO@SO3 CF3 ), and a sulfonated MWCNT (␮-CNT-SO3 H). To facilitate the separation of the catalysts from the reaction mixture MWCNTs were modified by incorporating magnetic nanoparticles. For comparison, the composites of MWCNTO with magnetic nanoparticles were further modified by deposition of niobia in two loadings, e.g. 30 and 60 wt.%. Depending of the functionality the catalysts exposed preponderantly either Brønsted (␮-CNT-SO3 H) or Lewis (GO@SO3 CF3 ) acid sites. Deposition of niobia also led to catalysts with acid properties, but as a function of the precursor nature and loading, they exposed niobia phases with either Brønsted (e.g., Nb-O units) or Lewis acid sites (e.g., −Nb = O units) in excess. Catalytic performances, in terms of the yields in one of the platform molecule (ie, levulinic acid versus lactic acid) were directly correlated with the nature of these acid sites. Interestingly enough, the catalysts showing high efficiency for lactic acid led also to high efficiency for the synthesis of succinic acid under pressure of molecular oxygen. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Research devoted to the chemicals, materials and ecologic fuels synthesis from renewable materials, as an alternative to the fossil resources, received a special attention in the past decades [1–3]. Part of this research was focused to the identification of new active catalytic systems adapted to the biomass particularities (e.g., the special nature of the bio-polymer raw material and low solubility in most of organic solvents) [2–5]. Former studies conducted in the cellulose and starch valorization, in the presence of homogeneous acid catalysts (19–35%, in HCl or H2 SO4 ), showed that the yields in levulinic acid (one of the most important platform molecules) are relatively low [6,7]. In addition, the use of mineral acids presented a series of drawbacks due to the corrosive nature of the catalyst, the difficulty of separation at the end of reaction and hazardous waste generated. Since the separation processes represent more than half of the total investment in

∗ Corresponding authors. E-mail addresses: [email protected] (V.I. Parvulescu), [email protected] (S.M. Coman).

equipment for the chemical and fuel industries, obviously industry prefer solid catalysts to homogeneous ones. However, these catalysts should be stable, should exhibit a water tolerant behavior and should be highly selective to guarantee the cost-effectiveness of the process [8]. In this context, many reports shown that grafted sulfonic groups on inorganic or organic polymers, as solid carriers, are hydrolyzed and leached into water at elevated temperatures [9]. Therefore, new active and selective solid catalyst systems are waiting to be discovered and developed. In recent years, niobia as well as modified niobia have been intensively investigated in a variety of important acid-catalyzed reactions, as dehydration, hydration, etherification, hydrolysis, and condensation, in which water molecules are directly involved as reactant or as reaction product [10–12]. One of the reasons of this high interest for niobia is its remarkable high acidity (Ho = − 8.2) which is also maintained in water [13–15]. Due to these characteristics, niobia-based catalysts were investigated by different research groups, in different transformations of biomass derivatives [16–19]. However, commercial bulk niobia suffers from poor hydrothermal stability resulting in a low catalytic stability under a high-water environment and temperatures higher than 200 ◦ C [20]. However, the deposition of niobia on oxide supports (e.g., silica, titania,

http://dx.doi.org/10.1016/j.cattod.2016.07.007 0920-5861/© 2016 Elsevier B.V. All rights reserved.

Please cite this article in press as: I. Podolean, et al., Efficient magnetic recoverable acid-functionalized-carbon catalysts for starch valorization to multiple bio-chemicals, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.07.007

G Model CATTOD-10316; No. of Pages 11

ARTICLE IN PRESS I. Podolean et al. / Catalysis Today xxx (2016) xxx–xxx

2

alumina) is somehow improving both the hydrothermal stability and the acidity in aqueous phase reactions [21]. Other types of catalysts that may potentially be considered for such processes are those incorporating triflic groups, like metal triflates or organic-triflate derivatives. To date, in spite of the remarkable acidity of triflic acid (it is one of the strongest Brønsted acids, with a pKa of −12 [22], of the very strong Lewis acidity of triflates, and of a very high tolerance towards water of both triflic acid and triflates, only scarce reports on their use in the direct valorization of biomass were published in literature [23]. Obviously, the functionalization of solid carriers with triflates is expected to provide environmentally friendly stable catalysts. However, there are also reports claiming that grafting with triflic acid is a matter of leaching if not stable bonds, on covalent nature, are formed between triflate species and carrier during impregnation process [24]. On the other side, nano-supported catalysts may bridge the gap between homogeneous and heterogeneous catalysis, preserving the desirable attributes of both systems [25]. In addition, the use of magnetically active nanoparticles like magnetite (Fe3 O4 ) may generate a more efficient separation of the catalyst simply by applying an external magnetic field. An example is the hydrolysis of cellulose to glucose with a solid acid Fe3 O4 –SBA–SO3 H with ordered mesopores and MNPs [26]. Likewise, not long ago, our group also demonstrated the efficiency of Fe3 O4 @SiO2 -Ru(III) catalysts, able to directly convert the cellulose into sorbitol and glycerol through three successive steps (ie, dehydration, formic acid decomposition, glucose hydrogenation with in situ produced hydrogen) [27] and levulinic acid into succinic acid through its oxidation with molecular oxygen [28]. In both cases the catalyst was easily recovered by simple applying a magnetic field on the external wall of the reactor. The developed nano-structured system also offers an excellent example on how the same catalyst can generate different active species able to produce multiple important platform molecules. Although not demonstrated, the SiO2 shell may be the subject of the hydrothermal degradation during the harsh reactions conditions. Based on this state of the art, in this study we explored novel heterogeneous catalytic systems, developed by functionalization of carriers with high mechanical strength properties (ie, graphene and MWCNT) with water tolerant acid species (ie, triflic acid, sulfonic acid and niobia). Very important, for an efficient separation of the catalyst, magnetic materials produced by incorporating superparamagnetic Fe3 O4 nanoparticles into remarkable mechanical stable MWCNTs were also investigated. The new designed catalysts were tested in starch valorization affording biomass-derived platform molecules like levulinic, lactic and succinic acids. 2. Experimental section All the chemicals and reagents were of analytical purity grade, purchased from Sigma-Aldrich and used without any further purification. Multiwalled carbon nanotubes (MWCNTs) were purchased from Sigma-Aldrich with following features: armchair configuration; preparation method Catalytic Chemical Vapor Deposition (CVD) (CoMoCAT® ), over 98% carbon, O.D. × L/6–9 nm × 5 ␮m. 2.1. Catalyst preparation Graphene-oxide (GO) and MWCNT-oxide (MWCNTO) were prepared by treating the pristine graphene and MWCNT with concentrated HNO3 following the next procedure: 250 mg of MWCNTs (or graphene) were dispersed by ultrasonication in 100 mL of HNO3 (1.0 mol/L) and vigorously stirred at 80 ◦ C for 6 h. After that, the separated solids were washed with distillated water until a neutral pH, and dried at 80 ◦ C for 24 h. Then the oxidized supports

were treated with i) concentrated triflic acid (10 mL acid/g support) under the same vigorous stirring at 80 ◦ C for 10 h, and ii) sulfuric acid (50 mL H2 SO4 98%) following the same protocol resulting in a sulfonated MWCNT/GO (␮-CNT-SO3 H and GO-SO3 H). However, the treatment of MWCNT with triflic acid failed. The graphene oxide functionalized with triflic acid was abbreviated as GO@SO3 CF3 . Further, magnetic nanoparticles were inserted in both the oxidized and sulfonated MWCNTs by using a method reported elsewhere [29]. Briefly, 1.8 g of Mohr’s salt (NH4 )2 Fe(SO4 )2 ·6H2 O, were dissolved in a mixture of 60 mL of degassed deionized water and hydrazine hydrate solution 50% wt in a ratio of 3:1, v/v. Then 0.5 g of MWCNTO or ␮-CNT-SO3 H was added in obtained grass-green solution, sonicated and stirred overnight under inert atmosphere and precipitated with ammonia solution (25 wt%) until the a pH = 13. The resulted precipitate was kept under stirring for another 3 h, at 80 ◦ C and then washed with water several times and separated by applying an external magnet. The resulted solids were denoted as ␮-CNT-OH (magnetic carbon nanotubes) and ␮-CNTSO3 H (sulphonated magnetic carbon nanotubes). ␮-CNT has also been modified by deposition of niobia in two loadings, e.g. 30 and 60 wt.%, on its surface. Two different precursors were used for this purpose: niobium ammonium oxalate C10 H5 NbO20 ·6H2 O (NAmOx) and niobium ethoxide Nb(OCH2 CH3 )5 (NbOEt). Thus, appropriate amounts of NAmOx or NbOEt were dissolved in ethanol that was followed by the subsequent addition of ␮-CNT. The resulted suspensions were maintained at 200 ◦ C for 48 h [30,31]. Then, the solids were separated with the aid of an external magnetic field, washed with distillated water, dried at 80 ◦ C and calcinated in air at 200 or 500 ◦ C, for 4 h (with a heating rate of 10 ◦ C/min). The ␮-CNT catalysts were denoted as ␮-CNT@A B C, where A is niobia precursor (NAmOx or NbOEt), B − niobia loading (30 or 60 wt%), and C – calcination temperature (200 or 500 ◦ C). Table 1 summarizes the composition of the final catalysts and the preparation parameters. 2.2. Catalyst characterization The obtained catalysts were exhaustively characterized using different techniques as powder X-ray diffraction (XRD), Raman and diffuse reflectance infrared Fourier transform (DRIFT) spectroscopy, adsorption-desorption isotherms of nitrogen at −196 ◦ C, ICP-OES, and elemental analysis. Powder X-ray diffraction (XRD) patterns were recorded using a Schimadzu XRD-7000 diffractometer with Cu K␣ radiation (␭ = 1.5418 Å, 40 kV, 40 mA) at a step of 0.2␪ and a scanning speed of 2 ◦ min−1 in the 5–90 ◦ 2␪ range. Raman spectra were collected with a Horiba Jobin Yvon − Labram HR UV–visible–NIR (200–1600 nm) Raman Microscope Spectrometer, using a laser with the wavelength of 633 nm. The spectra were collected from 10 scans at a resolution of 2 cm−1 . Diffuse reflectance infrared Fourier transform spectroscopy (DRIFT) spectra were collected with a Thermo 4700 spectrometer (200 scans with a resolution of 4 cm−1 ) on a domain of 600–4000 cm−1 . Surface areas were determined from the adsorption-desorption isotherms of nitrogen at −196 ◦ C using a Micromeritics ASAP 2020 Surface Area and Porosity Analyzer. To investigate the chemical stability of the catalysts the content of the leached metal into the reaction liquid was determined by ICP-OES (Agilent Technologies, 700 Series). Elemental analysis was performed using an EuroEA 3000 automated analyzer. The sample (less than 1 mg) was weighted in tin containers and was burned in a vertical reactor (oxidation tube) in the dynamic mode at 980 ◦ C in an He flow with the addition of O2 (10 mL) at the instant of sample introduction. Portions of the sample in tin capsules were placed in the automated sampler,

Please cite this article in press as: I. Podolean, et al., Efficient magnetic recoverable acid-functionalized-carbon catalysts for starch valorization to multiple bio-chemicals, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.07.007

G Model CATTOD-10316; No. of Pages 11

ARTICLE IN PRESS I. Podolean et al. / Catalysis Today xxx (2016) xxx–xxx

3

Table 1 Components of the catalysts and the preparation parameters. Entry

Catalyst abbreviation

Support

Catalytic phase precursor

Amount of catalytic phase, wt%

Drying/Calcination temperature,◦ C

1 2 3 4 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 4.10 4.11 4.12

GO@SO3 CF3 ␮-CNT-SO3 H GO-SO3 H ␮-CNT@A B C ␮-CNT@NAmOx 30 ␮-CNT@NAmOx 30 200 ␮-CNT@NAmOx 30 500 ␮-CNT @NAmOx 60 ␮-CNT@NAmOx 60 200 ␮-CNT@NAmOx 60 500 ␮-CNT@NbOEt 30 ␮-CNT@NbOEt 30 200 ␮-CNT@NbOEt 30 500 ␮-CNT@NbOEt 60 ␮-CNT@NbOEt 60 200 ␮-CNT@NbOEt 60 500

GO ␮-CNT GO ␮-CNT ␮-CNT ␮-CNT ␮-CNT ␮-CNT ␮-CNT ␮-CNT ␮-CNT ␮-CNT ␮-CNT ␮-CNT ␮-CNT ␮-CNT

Triflic acid Sulphuric acid Sulphuric acid A Niobium ammonium oxalate Niobium ammonium oxalate Niobium ammonium oxalate Niobium ammonium oxalate Niobium ammonium oxalate Niobium ammonium oxalate Niobium ethoxide Niobium ethoxide Niobium ethoxide Niobium ethoxide Niobium ethoxide Niobium ethoxide

10.31 (sulphur) 0.1 (sulphur) 0.2 (sulphur) B 30 30 30 60 60 60 30 30 30 60 60 60

80 80 80 C 80 200 500 80 200 500 80 200 500 80 200 500

from which were transferred to the oxidation tube at regular intervals. The concentration of each element was calculated using the Callidus program supplied with the analyzer. 2.3. Catalytic tests Prepared catalysts were tested in two different reactions: starch hydrolysis and oxidation of glucose and levulinic acid (LA) to succinic acid (SA). All tests were carried out in a glass-pressure autoclave (Top 45 model, Top Industrie). The activity tests were carried out in the following procedures: 2.3.1. Starch hydrolysis To a mixture of 50 mg of corn starch and 10 mL of water, 50 mg of catalyst (GO@SO3 CF3 , ␮-CNT-SO3 H and GO-SO3 H) were added and heated up to 130–180 ◦ C, under stirring (1.200 rpm), for 2–8 h. After that the catalyst was recovered (by filtration or by placing a permanent magnet on the reactor wall) and the water-soluble products were separated by distillation under vacuum. 2.3.2. Levulinic acid/glucose oxidation To a solution of 60 mg levulinic acid or glucose in 10 mL of water, 50 mg of ␮-CNT@A B C catalyst were added. After closing, the reactor was pressured at 10 bar with molecular oxygen and heated up to 180 ◦ C, under stirring (1200 rpm), for 6–24 h. After that, the oxygen was released and the catalyst was magnetically recovered by placing a permanent magnet on the reactor wall, and the products were separated by distillation under vacuum. 2.4. Products analysis Irrespective of the catalytic procedure, the recovered products were sylilated, dissolved in 1 mL of toluene and analyzed by GCFID chromatography (GC-Shimatzu apparatus). The identification of the water-soluble products was made using a GC–MS Carlo Erba Instruments QMD 1000 equipped with a Factor Four VF-5HT column with the following characteristics: 0.32 mm × 0.1 ␮m × 15 m working with a temperature program at a pressure of 0.38 Torr with He as the carrier gas. 3. Results and discussion In recent years, graphenes were extensively studied in the context of their multiple applications as, for instance, polymer composites, sensors, materials of “paper” transistors type, catalysts or biomedical applications. Owing to their large surface area, easy dispersability, and strong metal–support interactions, graphenes

Table 2 BET surface areas of the supports and the investigated catalysts. Catalyst

Surface area, m2 g−1

GO GO@SO3 CF3 ␮-CNT ␮-CNT-SO3 H

246 213 983 911

are also among the currently preferred supports. In this context, there are numerous examples showing that the catalytic activity of metallic-nanoparticles supported on graphene is higher than the activity of analogous metallic-nanoparticles supported on other carbon forms or metal oxides [32–34]. The oxidation of graphene to graphene oxide (GO) by treatment with strong mineral acids affords functionalization of basal planes and edges of the GO with exogenous groups, such as hydroxyl, epoxy, carbonyl or carboxyl [35]. In this way, the new functionalities may provide graphenes a high activity and selectivity in various reactions. On the other hand, it was already shown that the oxidation of CNT’s with mineral acids (e.g., nitric acid) begins at the tube ends, where the distribution of pentagons entails greatest lattice strain, leading to tip opened CNTs [36]. Accordingly, the acid treatment of CNTs also led to the structure’s erosion [37,38], and as the shortening and thinning of the CNT layers occurred, to the production of carbonaceous debris [39]. These phenomena easily explained the failure of the modification of the surface of CNT by triflic acid (pKa = −12) that is a much stronger acid than nitric (pKa = −1.3) or sulfuric acid (pKa1 = −3). The BET surface areas of the GO@SO3 CF3 , ␮-CNT-SO3 H catalysts and the carbonous carriers (ie, GO and ␮-CNT) are given in Table 2. The low differences between the values clearly indicate a highly dispersion of the functional groups onto the surface of the carriers, irrespective of their nature. Fig. 1 shows the XRD patterns of the graphene oxide (GO) and triflate-functionalized graphene oxide (GO@SO3 CF3 ). The XRD pattern of GO shows a strong diffraction line at 2␪ = 12.0◦ , which corresponds to an interlayer spacing of about 0.76 nm, indicating the presence of oxygen functionalities, which facilitate the hydration and exfoliation of GO sheets in aqueous media. After triflate chemical grafting the hydrophilicity of the water-dispersed GO sheets gradually decreased, leading to an irreversible agglomeration of GO sheets. Thus, the broad line centered at 2␪ 25◦ in the XRD pattern of the GO@SO3 CF3 sample confirms a random packing of the graphene sheets. The triflate groups are covalently bonded to the GO surface that is illustrated in the XRD

Please cite this article in press as: I. Podolean, et al., Efficient magnetic recoverable acid-functionalized-carbon catalysts for starch valorization to multiple bio-chemicals, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.07.007

G Model CATTOD-10316; No. of Pages 11 4

ARTICLE IN PRESS I. Podolean et al. / Catalysis Today xxx (2016) xxx–xxx

Fig. 1. XRD patterns of graphene oxide (GO) and triflate-functionalized graphene oxide (GO@SO3 CF3 ).

Fig. 2. DRIFT spectra of GO and GO@SO3 CF3.

pattern as a reduction of the oxidic groups and the removal of water molecules intercalated between the carbon planes. DRIFT spectrum of the GO and GO@SO3 CF3 samples also confirm the oxidation of the graphene and its functionalization with triflic acid (Fig. 2). The presence of different types of oxygen functionalities in GO were confirmed by the presence of the broad and wide band at 3447 cm−1 which can be attributed to the O H stretching vibrations of the C OH groups and water [40,41]. The absorption band at 1719 cm−1 can be attributed to C O carbonylic/carboxylic while the sharp intense peak 1636 cm−1 can be ascribed to benzene rings [42]. The absorption band at 1387 cm−1 can be attributed to C O carboxylic, the band at 1221 cm−1 to epoxy group and the band at 1055 cm−1 to C O from alkoxy groups.

Fig. 3. X-ray diffraction patterns of (a) CNT; (b) ␮-CNT-SO3 H.

The spectrum of GO@SO3 CF3 sample has been compared with the spectrum of GO. The grafting of triflic acid on GO gives rise to a new IR signal at 1234 cm−1 and two changes in the OH region. In particular it has been observed a decrease of the intensity of the 3447 cm−1 signal, due to vicinal C OH groups, and the formation of a broad absorption at 3200 cm−1 , due to H bonds between reactant and functional groups OH (Fig. 2). Obviously, the IR fingerprints of triflates are described by the bands at 1195 (stretching of CF3 group, very strong), 1400 (stretching of SO2 group, strong) and 3397 cm−1 (stretching of OH group, broad) [43]. Among these, only the CF3 vibration at 1234 cm−1 was well detectable for the GO@SO3 CF3 sample. However, there are also some indirect evidences. Thus, the significant decrease of the intensity of the bands attributed to −OH and C O groups after the triflic acid grafting, may account for the covalent binding of the triflate species to the GO surface, confirming the XRD results. Most probably, the functionalization of the GO surface to GO@SO3 CF3 takes place through a similar mechanism with that proposed by Perego and co-workers [24] for silica, as support. Two pathways are more probable: (i) triflic acid reacts with the C OH groups yielding highly dispersed hydrated triflic acid onto the GO surface (Scheme 1, structure 1), or (ii) triflic acid reacts with vicinal C OH groups yielding chemical bonds with the carbon structure. The water resulting from this esterification is then coordinated to the triflate leading to a grafted hydrated triflic acid (Scheme 1, structure 2). However, the XRD and DRIFT results gave more arguments in the favor of the structure 2 resulted from covalent bonds. The separation and recovery, usually performed by filtration or centrifugation, may degrade the catalysts after repeated reuses with a direct effect to the process effectiveness. The use of the magnetic materials may overcome these problems providing a more complete and simple catalyst recovery by means of an external magnetic field [27,28]. It may also expand the scope of these materials. Thus, carbon nanotubes filled with magnetic nanoparticles were reported as very interesting new materials for applications in biomedicine [29]. They have also successfully applied as carriers for the enzymes immobilization [44]. Based on these achievements we proposed to investigate sulfonated (␮-CNT-SO3 H) (Scheme 2) and niobia ␮-CNT@A B C-based magnetic carbon nanotubes as catalysts with potential efficiency in the biomass valorization (Scheme 3). Figs. 3 and 4 show the XRD patterns of these samples, where the reflections of the Fe3 O4 nanoparticles and carbon nanotubes are reliably identified.

Please cite this article in press as: I. Podolean, et al., Efficient magnetic recoverable acid-functionalized-carbon catalysts for starch valorization to multiple bio-chemicals, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.07.007

G Model CATTOD-10316; No. of Pages 11

ARTICLE IN PRESS I. Podolean et al. / Catalysis Today xxx (2016) xxx–xxx

Scheme 1. Possible structures of grafted triflic acid onto the GO surface in the synthesis of GO@SO3 CF3 sample.

Scheme 2. Schematic preparation of the sulfonated magnetic carbon nanotubes (␮-CNT-SO3 H).

Scheme 3. Schematic preparation of niobia ␮-CNT@A B C-based magnetic carbon nanotubes.

Please cite this article in press as: I. Podolean, et al., Efficient magnetic recoverable acid-functionalized-carbon catalysts for starch valorization to multiple bio-chemicals, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.07.007

5

G Model CATTOD-10316; No. of Pages 11 6

ARTICLE IN PRESS I. Podolean et al. / Catalysis Today xxx (2016) xxx–xxx

Intensity (a.u.)

d c

D G b a

2000 1800 1600 1400 1200 1000 800 600 400 200 -1

Raman Shift (cm )

Fig. 4. X-ray diffraction patterns of (a) CNT; (b) ␮-CNT-OH; (c) ␮-CNT @NAmOx 60; (d) ␮-CNT@NAmOx 60 200; (e) ␮-CNT@NAmOx 60 500; (f) ␮CNT@NAmOx 30 200; (g) ␮-CNT@NAmOx 30 500; (h) ␮-CNT @NbOEt 30 500 catalysts.

XRD pattern of CNT (Figs. 3 and 4, pattern a) indicates specific diffraction lines at 2␪ of around 26◦ and 44◦ (JCPDS file 65–6212) [45] assigned to (002) and (100) planes of the hexagonal graphite structure of MWCNTs. The XRD of pure Fe3 O4 nanoparticles (not shown in Figs. 3 and 4) showed diffraction lines at 2␪ 30.1◦ , 35.6◦ , 37.5◦ , 44.1◦ , 55.4◦ , 58.5◦ , 64.1◦ , and 74.6◦ assigned to (220), (311), (222), (400), (422), (511), (440), and (533) planes, corresponding to the cubic Fe3 O4 phase of the inverse spinel (JCPDS file 19-0629) [46]. No plane corresponding to oxidation of magnetite to oxides like hematite, were identified. From the XRD of MWCNTs filled with Fe3 O4 nanoparticles (shown as pattern b in Figs. 3 and 4), (220), (311), (222), (400), (422), (511), (440), and (533) planes of Fe3 O4 are also observed at 2␪ 30.06◦ , 35.45◦ , 37.00◦ , 42.99◦ , 53.50◦ , 57.04◦ , 62.56◦ , and 73.97◦ , respectively, in which the 2␪ values are shifted to lower angles than those found in pure Fe3 O4 due to the interaction of Fe3 O4 nanoparticles with MWCNTs. Moreover, the characteristic (002) and (100) diffraction planes lines of MWCNTs were observed at 2␪ of 26◦ and 44◦ , respectively. These results confirm a successful encapsulation of Fe3 O4 nanoparticles inside the MWCNTs nanotubes. The average crystallite size D of the encapsulated Fe3 O4 nanoparticles, calculated by using the Sheerer equation, is about 8 nm. The XRD pattern of the sulfonated magnetic carbon nanotubes (␮-CNT-SO3 H) (Fig. 3) is identical with that of MWCNTs filled with Fe3 O4 nanoparticles (shown as pattern b in Fig. 4) showing that the structure of the MWCNT is not damaged during the treatment with sulfuric acid. The deposition of niobia had a different effect than sulfonic acid. Thus, a slight modification in amorphous phase was observed for the C10 H5 NbO20 ·6H2 O (NAmOx) sample (pattern c from Fig. 4). Further calcination at 200 ◦ C resulted in the appearance of a new diffraction line at 2␪ = 22.5◦ (pattern d from Fig. 4), corresponding to a slight transition of niobia from amorphous to a crystalline phase. The increase of the calcination temperature at 500 ◦ C led to a new diffraction line at 2␪ 26.6◦ (pattern e from Fig. 4) that is attributed to the characteristic (400) plane of the NbO2 [47]. Interestingly enough, this transition occurred to much smaller temperatures (500 versus 800 ◦ C, typically) that evidences the role of the activated carbon assisting the Nb2 O5 phase transformation to NbO2 .

Fig. 5. Raman spectra of (a) ␮-CNT-OH; (b) ␮-CNT@NAmOx 30; (c) ␮-CNT @NAmOx 30 200; (d) ␮-CNT @NAmOx 30 500 catalysts.

Instead of NbO2 (pattern e from Fig. 4) the thermal treatment of NbOEt led to the formation of an orthorhombic crystalline Nb2 O5 phase (pattern h from Fig. 4) onto the CNT surface. Raising the calcination temperature over 500 ◦ C produces an irremediable damage of the magnetic nanoparticles. Starting from this temperature, dehydration of magnetite overpass 50%, leading to hematite (␣-Fe2 O3 ) as revealed from the diffraction lines at 2 = 33.1◦ , 40.1◦ , 49.3◦ , 53.9◦ , 62.5◦ , and 64.0◦ (JCPDS 86-0550) [48] (patterns g and h from Fig. 4). Raman spectra of the ␮-CNT carrier (Fig. 5(a) presented a series of sharp absorption lines in the range 100–700 cm−1 . According to the literature, the peaks located at ∼225, ∼290, ∼405, and ∼609 cm−1 are attributed to hematite (␣-Fe2 O3 ) [49,50]. This phase was also detected in XRD patterns but only for catalysts calcined at 500 ◦ C. Besides the presence of very small silent XRD particles, the presence of these lines might be also the consequence of a partial oxidation of magnetite to hematite by the laser radiation during the acquirement of the Raman spectra [51,52]. Raman spectra also presented the specific lines of multiwalled carbon nanotubes where the so called D line (∼1325 cm−1 ) corresponds to disordered carbon or defective graphitic structures, and the G line, at ∼1600 cm−1 , is characteristic to graphitic materials and arises from the stretching of the C C bond. The D to G intensity ratio (ID /IG ) indicates the degree of disorder in the carbon materials. For the ␮-CNT@NAmOx 30 catalysts (Fig. 5) the ID /IG ratio decreased with the increase of the calcination temperature till 200 ◦ C taking values of 1.79 for RT, 0.98 for 200 ◦ C, and 0.97 for 500 ◦ C, thus indicating a stabilization of the carbon structure [53]. For dried ␮-CNT@NAmOx 30 the broad Raman line centered at 640 cm−1 is ascribed to the main Nb O stretching mode of amorphous niobia (Fig. 5, spectrum b) while the lines located at 260 and 125 cm−1 were attributed to the lower vibrational modes of the NbO6 octahedra, suggesting angle deformation modes and bridging of Nb O Nb bonds. The calcination at 200 ◦ C led to a new line corresponding to Nb O stretching mode that shifts from 640 cm−1 to 660 cm−1 , indicating the transition from amorphous to TT-Nb2 O5 phase. This transition was also confirmed by low intensity liness in the region 800–1000 cm−1 . Similar lines were detected for the ␮CNT@NAmOx 30 catalyst calcined at 500 ◦ C. In addition this spectra presents lines associated to the presence of highly distorted NbO6 octahedra (800–1000 cm−1 ). Presumably, during the catalysts preparation niobium hydrated pentoxide Nb2 O5 x nH2 O is formed. The structure of this phase is not simple, exhibiting a multitude of polymorphic forms. The

Please cite this article in press as: I. Podolean, et al., Efficient magnetic recoverable acid-functionalized-carbon catalysts for starch valorization to multiple bio-chemicals, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.07.007

G Model CATTOD-10316; No. of Pages 11

ARTICLE IN PRESS I. Podolean et al. / Catalysis Today xxx (2016) xxx–xxx

formation of each is highly dependent on multiple factors like temperature, nature of the precursor, preparation route, etc. [54]. The spectra collected for the prepared catalyst indicated the presence of the amorphous and TT (orthorhombic structure) phases. The basic unit of amorphous niobia is represented by six-fold coordinated NbO6 distorted octahedral, while the building block of the TT-phase corresponds to a pseudo-hexagonal cell. Slightly distorted octahedra are constituted from Nb-O bonds and for a low temperature pretreatment are associated with the presence of the Brønsted acid sites. Calcination at 500 ◦ C makes possible the appearance of the highly distorted dehydrated NbO6 octahedra, associated with Lewis acid sites, due to the presence of the niobyl Nb = O entities [55]. The ICP-OES measurement of the spent catalysts indicated an insignificant leaching of Nb species. They were in a range of 1–3 ppm, while iron species were totally non-identified in the reaction medium. 3.1. Catalytic tests 3.1.1. Starch hydrolysis Acid hydrolysis of starch to produce sugars has been a commercial method since 1814, when the first plant to produce glucose syrup was build in France [56]. However, in addition to d-glucose, other products such as oligosaccharides, products of decomposition and dehydration of glucose as 5-hydroxymethylfurfural, levulinic acid, formic acid and colored products are formed as well [57]. At commercial level, enzymes are obviously used to increase the efficiency of the glucose production from starch and in order to retard the formation of reversion and decomposition products, starch is partially hydrolyzed with sulfuric liquid acid. After that, the hydrolysis is completed by enzymes [58]. However, both the use of mineral acids and fermentation by enzymes bring several disadvantages like large wastes production and reactor corrosion. Therefore the selective conversion of starch requires the development of catalytic materials with the ability to promote not only hydrolysis/isomerisation/hydration reactions but also to minimize the side reactions leading to by-products. Such requirements can be met by combining a certain type of active sites with a certain porosity of a solid catalytic support. Moreover, the use of the supports incorporating magnetic nanoparticles in association to highly dispersed active species (e.g., Lewis ( CF3 SO3 ) or Brønsted ( SO3 H) acid species) brought significant advantages over classical heterogeneous catalysts such as: (1) similar reaction rates to their homogeneous counterparts, (2) accessibility and high activity of the catalysts, (3) effective separation of the catalysts from the reaction medium by means of an external magnet. Fig. 6 displays the catalytic results obtained in the hydrolysis of starch in the presence of GO@SO3 CF3 and ␮-CNT-SO3 H catalysts. Similar catalytic results were obtained on both sulphuric acid treated carbon based carriers (ie, ␮-CNT-SO3 H and GO-SO3 H samples, not shown in Fig. 6). As Fig. 6 shows the catalysts act in a different way, producing different bio-chemicals (e.g., lactic or levulinic acid), with different efficiency. Unambiguously, the starch conversion in hot water is already a homogeneous Brønsted acid catalyzed process. However, water is not more acidic at 180 ◦ C than at 150 ◦ C and its higher ionic strength cannot explain alone the observed different reactivities. Therefore, as it was recently reported in cellulose hydrolysis a combination of different effects should be considered [59]. Thus, the important issue is to evaluate the role of the solid catalyst exhibiting Brønsted (␮-CNT-SO3 H) or Lewis (GO@SO3 CF3 ) acidity within the overall pathway initiated by homogeneous hydroxonium ions provided by the autoprotolysis of water at high temperatures. In the presence of GO@SO3 CF3 catalyst, the starch hydrolysis required a lower temperature (150 ◦ C, Fig. 6) while in the presence of ␮-CNT-SO3 H at 150 ◦ C humins represented the

7

Table 3 Elemental analysis of the GO@SO3 CF3 and ␮-CNT-SO3 H catalysts. Catalyst

N%

C%

H%

S%

␮-CNT-SO3 H ␮-CNT-SO3 Ha GO@SO3 CF3 GO@SO3 CF3 a

0.32 0.34 0.05 0.05

72.57 66.25 25.98 25.76

0.22 1.15 0.78 0.90

0.10 – 10.31 10.16

a

Spent catalyst.

only reaction product. The formation of water-soluble products was favored at temperatures above 180 ◦ C. Besides loading, this behavior clearly indicates a high difference in the strength of active sites. Elemental analysis shown that the loading of triflic sites immobilized onto GO was much higher than that of sulphonic acid immobilized onto the carbon nanotubes surface (10.31%S versus 0.10%S) (Table 3, entries 1 and 3). In the absence of any leaching, one can suppose a cooperative effect between the homogeneous acid catalysis promoted by hydroxonium ion from water autoprotolysis and the heterogeneous acid catalysis. The former one may be the driving force responsible of the starch hydrolysis into soluble compounds (glucose and its derivatives but also soluble oligomers and polymers, giving humins) while the later one would accelerate the transformation of glucose intermediate into final and stable products. In other words, this mixed homogeneous-heterogeneous acid catalysis appears to be rather negative as regard to the formation glucose. It is also important to notice that the stable product distribution was different as a function of the acid nature (e.g., Brønsted or Lewis). ␮-CNT-SO3 H catalyst, with Brønsted-type SO3 H sites preponderantly produced levulinic acid (S = 49%, 180 ◦ C, 6 h). This high selectivity in levulinic acid can be explained by accepting that the oxygen-containing groups from the surface of the functionalized ␮-CNT are not spectators and play a key role in the adsorption and the transferring of the reactants [60]. The formation of levulinic acid involves several key steps, including the transformation of starch to HMF, and the subsequent rehydration of HMF to levulinic acid (Scheme 4, path (B)). The ␮CNT-SO3 H Brønsted acid catalyst plays a key role in the conversion of HMF into levulinic and formic acids. However, although high yields of levulinic acid can be attained by using the ␮-CNT-SO3 H catalyst, we found that repeated uses of this catalyst are limited due to the leaching of acidic groups (Table 3, entry 2). Therefore, the picture of the process is, in this case, even more complicate involving autoprotolysis/homogeneous catalysis and heterogeneous catalysis. A quite different picture was observed for the GO@SO3 CF3 Lewis acid catalyst. Its use not only enhanced the conversion of starch, but also switched the selectivity from levulinic to lactic acid (Scheme 4, path (A)). In its presence, lactic acid was produced with a selectivity of 62.9% at 150 ◦ C in only 4 h of reaction. Most probably, such a behavior is associated to the fact that in the presence of the ␮-CNT-SO3 H catalyst the energy barrier for dehydration (the pathway involving levulinic acid formation) was lower than that for the retro-aldol fragmentation (the pathway involving lactic acid formation). On the other hand, the participation of the GO@SO3 CF3 Lewis catalyst decreased the energy barrier for the retro-aldol reaction, thus leading to the shift in the selectivity from levulinic to lactic acid. It is also important to notice that the reaction pathway in the acid-catalyzed conversion of starch is highly sensitive to the type of acid. While ␮-CNT-SO3 H (Brønsted acid) catalyzes the dehydration of glucose, leading to the formation of levulinic acid, GO@SO3 CF3 (Lewis acid) catalyzes the retro-aldol reaction, resulting in the formation of ␣-hydroxycarboxylic acid derivatives (e.g., lactic and glycolic acid) (Scheme 4). It also appeared from these experiments

Please cite this article in press as: I. Podolean, et al., Efficient magnetic recoverable acid-functionalized-carbon catalysts for starch valorization to multiple bio-chemicals, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.07.007

G Model CATTOD-10316; No. of Pages 11 8

ARTICLE IN PRESS I. Podolean et al. / Catalysis Today xxx (2016) xxx–xxx

Fig. 6. Starch hydrolysis in the presence of the GO@SO3 CF3 and ␮-CNT-SO3 H catalysts (conversions were 100% irrespective of the catalyst nature or reaction conditions. Reaction conditions: 50 mg starch, 10 mL water, 50 mg catalyst (GO@SO3 CF3 and ␮-CNT-SO3 H)).

Scheme 4. Catalytic pathway using the GO@SO3 CF3 Lewis acid catalyst.

that functionalized carbon carriers (GO or CNT) are active for the isomerization of glucose to fructose. It also results that to achieve a high selectivity to lactic acid it is important to suppress the catalytic effect of Brønsted acid sites, as in the case of GO@SO3 CF3 sample. The obtained results confirm the already reported findings in the chemocatalytic conversion of cellulose in water in the presence of dilute lead(II) ions and extend them to heterogeneous catalysis [61].

3.1.2. Levulinic acid/glucose oxidation In the investigated conditions the oxidation of levulinic acid (LA, Scheme 5) occurred with the complete conversion and selectivities to succinic acid (SA) higher than 98% on ␮-CNT@NbOEt 30 200 and ␮-CNT@NbOEt 30 500 catalysts (Fig. 7). Using glucose as raw material (Scheme 6), the level of SA selectivity was also quite high (around 80%) for a total conversion of glucose (Fig. 7). In this case,

Please cite this article in press as: I. Podolean, et al., Efficient magnetic recoverable acid-functionalized-carbon catalysts for starch valorization to multiple bio-chemicals, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.07.007

G Model

ARTICLE IN PRESS

CATTOD-10316; No. of Pages 11

I. Podolean et al. / Catalysis Today xxx (2016) xxx–xxx

9

100% 90% 80% 70% 60%

%

Others FDCA 3 HP Furoic acid Lactic acid Succinic acid

50% 40% 30% 20% 10% 0% LA

G

S

LA

µ-CNT@NbOEt_30_200

G

S

µ-CNT@NbOEt_30_500

Raw material/catalyst Fig. 7. Oxidation of LA-levulinic acid, G-glucose, S-starch on the ␮-CNT@NbOEt 30 200 and ␮-CNT@NbOEt 30 500 catalysts (Reaction conditions: 60 mg substrate, 10 mL water, 50 mg ␮-CNT@A B C catalyst, 10 bar O2 , 180 ◦ C, 6 h in the case of LA and G and 24 h in the case of S).

O OH

H3C

O

catalyst + 3/2 O2

OH

HO

H

O

O

OH

(HCO2H, formic acid)

(C4H6O4, succinic acid)

(C5H8O3, levulinic acid)

O

+

Scheme 5. Oxidation of levulinic acid to succinic acid.

OH HO HO

O

HO OH

OH

HOOC H+

H C

H C COOH

OH OH

H

catalyst

HOOC

C C COOH H2 H2

succinic acid Scheme 6. Oxidation of glucose to succinic acid.

the main by-products were the furoic, lactic, 3-hydroxypropanoic (3HP) and 2,5-furandicarboxylic (FDCA) acids. When starch was used as substrate, the selectivity to succinic acid was highly decreased (around 20% for a conversion of 50%). Fig. 7 presents the influence of the calcination temperature on the catalysts performances. In all cases the calcination of the catalysts at 500 ◦ C resulted in a slight increase of the selectivity to

succinic acid. The highest increase occurred when used glucose as raw material. In this case, the calcination of the catalyst at 500 ◦ C resulted in an increased selectivity to succinic acid with 12%. The calcination also affects the magnetite phase leading to the formation of hematite. Hematite is not a superparamagnetic structure that may affect the simple separation of the catalysts, but is not influencing the catalytic performances of the samples. Fortunately, as XRD patterns shown (see Fig. 4), only a small fraction of magnetite was transformed into hematite. Thus the separation of the catalysts by magnetic forces was not affected. As it was shown above, the Raman spectra indicated the presence of the amorphous (with a slightly distorted octahedral unit of six-fold coordinated NbO6 ) and TT (orthorhombic structure with a pseudo-hexagonal cell) phases. Slightly distorted octahedra are associated with the presence of the Brønsted acid sites (catalysts calcined at 200 ◦ C), while the highly distorted dehydrated NbO6 octahedra are associated to Lewis acid sites due to the presence of the niobyl Nb O entities (calcination at 500 ◦ C). From the results we collected it results that the presence of niobyl entities is responsible to the slight increase of the selectivity in succinic acid. However, as XRD shown, the catalysts calcined at 500 ◦ C lose quickly their paramagnetic behavior, due to the conversion of magnetite to hematite. This makes quite difficult the catalysts separation and recycling from the reaction media. Therefore, from practical reasons it is preferable the calcination of catalysts at lower temperatures, as long as the differences in the selectivity to succinic acid are not notable. The effect of niobium loading and the nature of its precursor were also studied in the glucose oxidation (Fig. 8). ␮-CNT@NAmOx 60 200 catalyst, with 60 wt% niobium prepared from niobium ammonium oxalate (NAmOx), led to very poor selectivities in SA (SSA = 3.0%) as compared to the ␮-CNT@NbOEt 30 200 catalyst, with 30 wt% niobium (SSA = 70.3%). Interestingly enough,

Please cite this article in press as: I. Podolean, et al., Efficient magnetic recoverable acid-functionalized-carbon catalysts for starch valorization to multiple bio-chemicals, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.07.007

G Model

ARTICLE IN PRESS

CATTOD-10316; No. of Pages 11

I. Podolean et al. / Catalysis Today xxx (2016) xxx–xxx

10

100% 90% 80% 70% 60%

%

50% 40%

Others FDCA Furoic acid 3HP Lactic acid Succinic acid

30% 20% 10% 0% µ-CNT@NbOEt_30_200

µ-CNT@NAmOx_60_200

µ-CNT@NAmOx_30_200

Catalyst Fig. 8. Glucose oxidation with different niobium-based catalysts (Reaction conditions: 50 mg glucose, 10 mL water, 50 mg ␮-CNT@A B C catalyst, 10 bar O2 , 180 ◦ C, 6 h).

Table 4 Niobium-based catalysts performances in the hydrolysis and oxidation of glucose. Entry

Catalyst

PO2 , atm

C (%)

1 2 3 4

␮-CNT@NbOEt 30 200 ␮-CNT@NbOEt 30 200 ␮-CNT@NAmOx 30 200 ␮-CNT@NAmOx 30 200

– 10 – 10

100 100 98.2 100

S (%) LA

SA

51.7 6.3 26.4 4.9

– 70.3 – 45.9

lower loadings of niobium (NAmOx) (i.e., 30 wt%) led to much higher selectivities in SA (SSA = 45.8%). As characterization of the catalysts shown, different niobiumbased phases are formed as a function of the nature of niobia precursors. Niobium ammonium oxalate (NAmOx) led to agglomerations of NbO2 phase while niobium ethoxide (NbOEt) led to an orthorhombic crystalline Nb2 O5 phase. However, below 200 ◦ C these crystalline phases are not completely formed (see XRD patterns, Fig. 4) and the dominant phase is the amorphous one. On the other hand, the increase of the niobium loading from 30 to 60 wt% led to larger agglomerations of Nb(IV)/Nb(V)-phases. Differences in these agglomerations were also observed. For NAmOx they correspond to crystallites rich in NbO2 with a dominant Brønsted acid character, while for NbOEt to a Nb2 O5 crystalline phase, with preponderantly Lewis acid character where the presence of the Nb O bonds seems to favor the oxidation to succinic acid. As in the case of GO@SO3 CF3 (Lewis acid), in the hydrolysis of glucose, the presence of niobium-based catalysts with preponderantly Lewis sites, namely ␮-CNT@NbOEt samples led to lactic acid in high selectivities (Table 4). The achieved data indicate that although Brønsted acidity favors the hydrolysis of glucose, an excess in these species does not help

a further transformation of glucose to lactic acid. In both cases, a high amount of species with Lewis acid character (irrespective is a triflate or niobium) seems to adjust the acid Lewis/Brønsted ratio to one favorable to lactic acid synthesis [19].

4. Conclusions The catalytic transformation of starch into platform or buildingblock chemicals is promising for the utilization of the abundant agriculture and municipal wastes. The present study proposes efficient catalysts for the synthesis of levulinic, lactic and succinic acids. The formation of levulinic acid involves a series of tandem reactions including the hydrolysis of starch, the isomerization of glucose, dehydration of fructose and rehydration of HMF. Brønsted acid catalysts, such as ␮-CNT-SO3 H, are capable to catalyze each step of these tandem reactions via a complex autoprotolysis/homogeneous/heterogeneous process, but the repeated uses of this catalyst is limited due to the leaching of acidic groups under the harsh reaction conditions. Lewis acids, in particular triflate-based graphene oxide (GO@SO3 CF3 ) and niobium-based catalysts immobilized on supports incorporating magnetic nanoparticles (␮-CNT@NbOEt) led to the formation of lactic acid in water. The pathway involved the following tandem steps: hydrolysis of starch to glucose, isomerization of glucose to fructose, retro-aldol fragmentation of fructose to C3 intermediates and the isomerization of the trioses into lactic acid. Adjusting the preparation conditions of the niobium based catalysts (e.g., precursor nature, calcination temperature and amount of niobium on the ␮-CNT surface) led to efficient catalysts for the synthesis of lactic acid. Interestingly enough the same catalysts are highly active for the levulinic acid or glucose oxidation to succinic acid. Moreover, the catalysts can be easily separated and recycled several times by applying an external magnet on the reactor wall.

Please cite this article in press as: I. Podolean, et al., Efficient magnetic recoverable acid-functionalized-carbon catalysts for starch valorization to multiple bio-chemicals, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.07.007

G Model CATTOD-10316; No. of Pages 11

ARTICLE IN PRESS I. Podolean et al. / Catalysis Today xxx (2016) xxx–xxx

Acknowledgments The work was supported by the strategic grant POSDRU/159/1.5/S/137750, Project “Postdoctoral programme for training scientific researchers” co-financed by the European Social Foundation within the Sectorial Operational Program Human Resources Development 2007–2013. Prof. Simona M. Coman kindly acknowledges UEFISCDI for the financial support (project PN-IIPCCA-2011-3.2-1367, Nr. 31/2012). References [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]

P. Gallezot, Chem. Soc. Rev. 41 (2012) 1538–1558. G. Dautzenberg, M. Gerhardt, B. Kamm, Holzforschung 65 (2011) 439–451. A.T.W.M. Hendriks, G. Zeeman, Bioresour. Technol. 100 (2009) 10–18. P. Alvira, E. Tomás-Pejó, M. Ballesteros, M.J. Negro, Bioresour. Technol. 101 (2010) 4851–4861. P. Kumar, D.M. Barett, M.J. Delwiche, P. Stroeve, Ind. Eng. Chem. Res. 48 (2009) 3713–3729. W.W. Moyer, Preparation of levulinic acid. US patent 2270328. C.H.G. Hands, F.R. Whitt, J. Soc. Chem. Ind. 66 (1947) 415–416. R. Rinaldi, F. Schüth, Energy Environ. Sci. 2 (2009) 610–626. A. Onda, T. Ochi, K. Yanagisawa, Green Chem. 10 (2008) 1033–1037. V. Pârvulescu, M. Ruwet, P. Grange, V.I. Pârvulescu, J. Mol. Catal. 135 (1998) 75–88. L.C.A. Oliveira, T.C. Ramalho, E.F. Souza, M. Gonc¸alves, D.Q.L. Oliveira, M.C. Pereira, J.D. Fabris, Appl. Catal. B Environ. 83 (2008) 169–176. A.C. Silva, D.Q.L. Oliveira, L.C.A. Oliveira, A.S. Anastácio, T.C. Ramalho, J.H. Lopes, H.W.P. Carvalho, C.E.R. Torres, Appl. Catal. A Gen. 357 (2009) 79–84. T. Okuhara, Chem. Rev. 102 (2002) 3641–3666. G. Busca, Chem. Rev. 107 (2007) 5366–5410. S. Suganuma, K. Nakajima, M. Kitano, D. Yamaguchi, H. Kato, S. Hayashi, M. Hara, J. Am. Chem. Soc. 130 (2008) 12787–12793. D. Ding, J. Wang, J. Xi, X. Liu, G. Lu, Y. Wang, Green Chem. 16 (2014) 3846–3853. P. Carniti, A. Gervasini, M. Marzo, Catal. Commun. 12 (2011) 1122–1126. C. García-Sancho, I. Sádaba, R. Moreno-Tost, J. Mérida-Robles, J. Santamaría-González, M. López-Granados, P. Maireles-Torres, ChemSusChem 6 (2013) 635–642. S.M. Coman, V. Verziu, A. Tirsoaga, B. Jurca, C. Teodorescu, V. Kuncser, V.I. Parvulescu, G. Scholz, E. Kemnitz, ACS Catalysis 5 (2015) 3013–3026. Y.J. Pagán-Torres, J.M.R. Gallo, D. Wang, H.N. Pham, J.A. Libera, C.L. Marshall, J.W. Elam, A.K. Datye, J.A. Dumesic, ACS Catal. 1 (2011) 1234–1245. M. Ziolek, Catal. Today 78 (2003) 47–64. E. Raamat, K. Kaupmees, G. Ovsjannikov, A. Trummal, A. Kütt, J. Saame, I. Koppel, I. Kaljurand, L. Lipping, T. Rodima, V. Pihl, I.A. Koppel, I. Leito, J. Phys. Org. Chem. 26 (2013) 162–170. P.J. Deuss, K. Barta, J.G. de Vries, Catal. Sci. Technol. 4 (2014) 1174–1196. A. de Angelis, C. Flego, P. Ingallina, L. Montanari, M.G. Clerici, C. Carati, C. Perego, Catal. Today 65 (2001) 363–371. H. Frei, J.Y. Park, J. Am. Chem. Soc. 131 (2009) 16589–17032. D.-m. Lai, L. Deng, J. Li, B. Liao, Q.-x. Guo, Y. Fu, ChemSusChem 4 (2011) 55–58. A. Negoi, I.T. Trotus, O. Mamula Steiner, M. Tudorache, V. Kuncser, D. Macovei, V.I. Parvulescu, S.M. Coman, ChemSusChem 6 (11) (2013) 2090–2094. I. Podolean, V. Kuncser, N. Gheorghe, D. Macovei, V.I. Parvulescu, S.M. Coman, Green Chem. 15 (2013) 3077–3082.

11

[29] J.H. Li, R.Y. Hong, G.H. Luo, Y. Zheng, H.Z. Li, D.G. Wei, New Carbon Mater. 25 (2010) 192–198. [30] W. Chen, X. Cui, Production of decorated carbon nanotubes, WO2011143777 A1, 2011. [31] S. Damyanova, L. Dimitrova, L. Petrova, P. Grange, Appl. Surf. Sci. 214 (2003) 68–74. [32] X. Chen, G. Wu, J. Chen, X. Chen, Z. Xie, X. Wang, J. Am. Chem. Soc. 133 (2011) 3693–3695. [33] A. Primo, I. Esteve-Adell, J.F. Blandez, A. Dhakshinamoorthy, M. Alvaro, N. Candu, S.M. Coman, V.I. Parvulescu, H. Garcia, Nat. Commun. (2015), http:// dx.doi.org/10.1038/ncomms9561. [34] A. Primo, I. Esteve-Adell, S.N. Coman, N. Candu, V.I. Parvulescu, H. Garcia, Angew. Chem. Int. Ed. (2015), http://dx.doi.org/10.1002/anie.201508908. [35] A. Lerf, H.Y. He, M. Forster, J. Klinowski, J. Phys. Chem. B 102 (1998) 4477–4482. [36] H. Hiura, T.W. Ebbesen, K. Tanigaki, Adv. Mater. 7 (3) (1995) 275–276. [37] M.Q. Tran, C. Tridech, A. Alfrey, A. Bismarck, M.S.P. Shaffer, Carbon 45 (2007) 2341–2350. [38] R. Verdejo, S. Lamoriniere, B. Cottam, A. Bismarck, M. Shaffer, Chem. Commun. (2007) 513–515. [39] H. Hu, B. Zhao, M.E. Itkis, R.C. Haddon, J. Phys. Chem. B 107 (50) (2003) 13838–13842. [40] S. Bykkam, K.V. Rao, C.H.S. Chakra, T. Thunugunta, Int. J. Adv. Biotechnol. Res. 4 (2013) 142–146. [41] T. Zhang, D. Zhang, M.A. Shen, J. Mater. Lett. 63 (2009) 2051–2054. [42] Y. Si, E.T. Samulski, J. Nano Lett. 8 (2008) 1679–1682. [43] S.M. Coman, G. Pop, C. Stere, V.I. Parvulescu, J. El Haskouri, D. Beltrán, P. Amorós, J. Catal. 251 (2007) 388–399. [44] W.J. Goh, V.S. Makam, J. Hu, L. Kang, M. Zheng, S.L. Yoong, C.N.B. Udalagama, G. Pastorin, Langmuir 28 (2012) 16864–16873. [45] JCPDS, 2001. International Centre for Diffraction Data 25-0284. [46] Y. Chen, H. Gu, Mater. Lett. 67 (2012) 49–51. [47] K. Huang, Y. Li, L. Yan, Y. Xing, RSC Adv. 4 (2014) 9701–9708. [48] M. Yamaura, R.L. Camilo, L.C. Sampaio, M.A. Macedo, M. Nakamura, H.E. Tomas, J. Magn. Magn. Mat. 279 (2004) 210–217. [49] P. Lottichi, C. Barato, D. Bersani, G. Antonioli, A. Montenero, M. Guarneri, Opt. Mater. 9 (1998) 368–373. [50] A.P. Kozlova, S. Sugiyama, A.I. Kozlov, K. Asakura, Y. Iwasawa, J. Catal. 176 (1998) 426–438. [51] I. Chourpa, L. Douziech-Eyrolles, L. Ngaboni-Okassa, J.F. Fouquenet, S. Cohan-Jonathan, M. Souce, H. Marchais, P. Dubois, Analyst 130 (2005) 1395–1403. [52] O.N. Shebanova, P. Lazor, J. Solid State Chem. 174 (2003) 424–430. [53] S. Osswald, Y. Gogotsi, In situ Raman spectroscopy of oxidation of carbon nanomaterials, in: C.S.S.R. Kumar (Ed.), Raman Spectroscopy for Nanomaterials Characterization, Springer, Berlin, 2012, pp. 291–351. [54] H. Schafer, R. Gruehn, F. Schulte, Angew. Chem. Int. Ed. Engl. 5 (1966) 40–52. [55] I. Nowak, M. Ziolek, Chem. Rev. 99 (1999) 3603–3624. [56] G. Reed, Enzyme in Food Processing, Academic Press, New York, 1966. [57] R.W. Kerr, F.C. Cleveland, Starch 5 (1953) 261–266. [58] W.J. Whelan, Starch 9 (1957) 74–76. [59] M. Möller, P. Nilges, F. Harnisch, U. Schröder, ChemSusChem 23 (2011) 566–579. [60] P.P. Upare, J. Yoon, M.Y. Kim, H.Y. Kang, D.W. Hwang, Y.K. Hwang, H.H. Kung, J.S. Chang, Green Chem. 15 (2013) 2935–2943. [61] Y. Wang, W. Deng, B. Wang, Q. Zhang, X. Wan, Z. Tang, Y. Wang, C. Zhu, Z. Cao, G. Wang, H. Wan, Nat. Commun. 4 (2013) 2141, http://dx.doi.org/10.1038/ ncomms3141.

Please cite this article in press as: I. Podolean, et al., Efficient magnetic recoverable acid-functionalized-carbon catalysts for starch valorization to multiple bio-chemicals, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.07.007