Porous structured CuO-CeO2 nanospheres for the direct oxidation of cellobiose and glucose to gluconic acid

G Model

ARTICLE IN PRESS

CATTOD-10547; No. of Pages 11

Catalysis Today xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

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

Porous structured CuO-CeO2 nanospheres for the direct oxidation of cellobiose and glucose to gluconic acid Prince Nana Amaniampong a,1 , Quang Thang Trinh b , Kaixin Li a , Samir H. Mushrif a , Yu Hao c,∗ , Yanhui Yang a,∗ a

School of Chemical and Biomedical Engineering, Nanyang Technological University, Singapore 637459, Singapore Cambridge Centre for Advanced Research and Education in Singapore (CARES), Nanyang Technological University, 1 Create Way, Singapore 138602, Singapore c School of Chemistry and Chemical Engineering, South China University of Technology, Guangzhou 510640, China b

a r t i c l e

i n f o

Article history: Received 27 September 2016 Received in revised form 30 November 2016 Accepted 3 January 2017 Available online xxx Keywords: C H bond activation Lattice oxygen Nanospheres Gluconic acid Oxidation

a b s t r a c t Porous-structured CuO-CeO2 nanospheres were synthesized using a hydrothermal method and were tested as catalysts for the direct oxidation of cellobiose to gluconic acid. Catalytic reaction along with catalyst characterization results and 18 O-oxygen isotope labeled experiments revealed that the surface lattice oxygen of CuO in CuO-CeO2 nanospheres was consumed during the oxidation of cellobiose. This provides a direct evidence of our previous work (Amaniampong et al., Angew. Chem. Int. Ed. 54 (2015) 8928–8933). Characterization results further suggested that the lattice oxygen in CeO2 did not participate in the oxidation; nonetheless, the addition of CeO2 to CuO enhanced the surface area of the catalyst composite which was crucial for the reaction. The spent catalyst upon re-oxidation regained its activity. In addition, isotope labeled deuterium oxide (D2 O) experiments suggested that hydrogen exchange between the solvent and the substrate (glucose) are not involved in the mechanistic formation of gluconic acid and confirmed the solvent had no direct influence in the formation of gluconic acid. © 2017 Elsevier B.V. All rights reserved.

1. Introduction Metal oxide catalysts are promising for oxidation processes, due to their lower costs and higher stability, as compared to noble metals. The synthesis and characterization of transition metal oxides with controlled shape, size, structure and uniform dimension have attracted considerable interest because of their unique physicochemical properties and potential in catalysis [1,2]. Among these, copper oxide (CuO) was reported to be suitable for various catalytic applications [3–6], and the property of CuO has been documented in literature [7] to be closely related to its microstructures, particularly its morphology, crystal size and orientation. Rajesh et al. reported that CuO/Al2 O3 was even more active than Pt/Al2 O3 for the complete oxidation of ethanol [8]. Larsson and Anderson attempted CuOX /Al2 O3 for the incineration of CO, ethanol and ethyl acetate and found an excellent catalytic performance [9,10]. CuO was also

∗ Corresponding authors. E-mail addresses: [email protected] (Y. Hao), [email protected] (Y. Yang). 1 Current address: INCREASE (FR CNRS 3707), ENSIP, 1 rue Marcel Doré, TSA41105, 86073, Poitiers Cedex 9, France.

the most active metal oxide of those explored for the catalytic incineration of toluene with ␥-Al2 O3 as support [11]. Copper-modified or copper promoted by ceria has been reported to exhibit enhanced catalytic performance than copper only in the total oxidation of carbon monoxide, catalytic wet oxidation of phenol and water gas shift reactions [12–14]. The enhanced activity of the copper oxide promoted by cerium oxide can be attributed to the structural enhancement and improved thermal stability of the copper-cerium mixed oxide [15]. The activation of C H bond was reported to be a key step in oxidation reactions [16,17]. There has been considerable interest in the oxidation of saturated hydrocarbons involving C H bond activation to manufacture valuable products such as the production of oxygenates via the partial oxidation of light alkanes [18] through the activation of C H bonds by under coordinated metal lattice oxygen sites, with a concerted mechanism in which the metal center inserted into the C H bond, and cleaved it. The production of olefins via dehydrogenation or oxidative dehydrogenation mechanisms over transition metal oxides has also become attractive [19]. Recently, great interest in the valorization of renewable resources has led to extensive investigation on the catalytic oxi-

http://dx.doi.org/10.1016/j.cattod.2017.01.009 0920-5861/© 2017 Elsevier B.V. All rights reserved.

Please cite this article in press as: P.N. Amaniampong, et al., Porous structured CuO-CeO2 nanospheres for the direct oxidation of cellobiose and glucose to gluconic acid, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2017.01.009

G Model CATTOD-10547; No. of Pages 11

ARTICLE IN PRESS P.N. Amaniampong et al. / Catalysis Today xxx (2016) xxx–xxx

2

dation of sugars and other carbohydrates. In spite of the great interest in these processes, partially conflicting opinions regarding their underlying mechanisms, were reported in the literature concerning several details of the mechanism of oxidation over oxide catalysts. For instance, the selective oxidation of glucose to gluconic acid, and the liquid phase oxidation at the alcoholic C OH and carbonylic C O bonds have been reported to occur via dehydrogenation mechanism [20,21]. In our previous study [6], oxidation of glucose, cellobiose and cellulose was performed on a CuO catalyst, in the form of nanoleaves, with excellent yield towards gluconic acid. Using quantum mechanical calculations, it was revealed that the surface lattice oxygen of CuO activated the formyl C H bond in glucose and incorporated itself into the glucose molecule to oxidize it to gluconic acid. Herein, CuO-CeO2 nanospheres with a porous structure were synthesized by a hydrothermal treatment, using urea as a precipitating agent without the aid of surfactant to achieve a Ce-Cu binary precursor and subsequent calcination of the precursor. The assynthesized CuO-CeO2 nanospheres were characterized by X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), highresolution transmission electron microscope (HR-TEM), transmission electron microscope (TEM), H2 temperature-programmed reduction (H2 -TPR), N2 adsorption-desorption techniques and scanning electron microscope (SEM). The catalytic performance for the selective oxidation of cellobiose and glucose was examined to study the correlation between the catalytic performance and the interaction at the Ce-Cu interface. It was anticipated that, high oxygen mobility and redox properties of ceria oxide my further promote the catalytic properties of copper oxide in these reactions due to the additional active sites generated from oxygen vacancies at the Cu-Ce interfaces. Isotope labeling experiments with 18 O (oxygen) were carried out to study the reaction mechanism and also to further confirm the findings in our previous study that, indeed it was the lattice oxygen in CuO that incorporated into glucose to form gluconic acid [6]. Deuterium labeled water (D2 O) experiments were also carried out to investigate the role of solvent.

equipped with TCD. In each run, approximately 50 mg of the catalyst was pretreated at 300 ◦ C under a flow of He (30 mL min−1 ), and then heated to 700 ◦ C with a ramp of 10 ◦ C min−1 in the stream of 5 vol% H2 /Ar (40 mL min−1 ). Surface area analysis was determined by nitrogen physisorption on a Micromeritics TrisStar apparatus. The specific area was calculated using the Brunauer-Emmett-Teller (BET) equation. Raman measurements were performed on an inVia reflex confocal microprobe Raman system (Renishaw Company). Excitation with radiation of 514.5 nm was provided with an Ar+ laser. 2.2. Catalyst testing Catalytic experiments were carried out using 0.050 g of CuOCeO2 catalyst and 0.205 g of cellobiose, unless otherwise stated. A stainless-steel autoclave reactor, equipped with a Teflon liner (50 mL) was employed to perform the oxidation of cellobiose. Typically, cellobiose, the catalyst and deionized water (15 mL) were loaded into the reactor. The reactor was purged several times with high-purity nitrogen, (unless otherwise stated) to eliminate any traces of residue air present in the reactor. Reactions were allowed to proceed at desired set temperatures with constant stirring at 800 rpm. After the reaction was completed, an Agilent 1100 HPLC with a RID − 6A refractive index detector and a Hi-Plex H column (300 × 6.5 mm) were used to analyze the reaction products, with a mobile phase of 0.01 M H2 SO4 buffer at a 1 mL min−1 flow rate. For the isotopically labeled experiments, D2 O was used as solvent and O18 -labeled oxygen was used for the spent catalyst regeneration. Noteworthy, the detected amount of formic acid reported in our study is the solubilized fraction of it, although the volatized amount of formic acid at the analyze condition is small according to its corresponding Henry’s law coefficient [23]. 3. Results 3.1. Structural characterization

2. Experimental 2.1. Catalyst preparation and characterization CuO-CeO2 nanospheres were prepared by a hydrothermal method based on a previously reported two-step route [22]. In a typical synthesis process, Ce(NO3 )3 ·6H2 O (1 mmol) and Cu(NO3 )2 ·3H2 O (1 mmol) were dissolved in 50 mL of deionized water, followed by adding 8 mmol of urea dropwise. The transparent solution attained after stirring the precursor solution, was transferred to a Teflon-lined autoclave and then heated at 180 ◦ C for 100 min, resulting in the precipitation of the precursors. The precipitate was centrifuged, washed several times with deionized water and ethanol. After drying at ambient atmosphere, the precursor was finally calcined in air at 600 ◦ C for 4 h. For the spent catalyst regeneration, a specified amount of the CuO-CeO2 catalyst recovered after reaction were thoroughly washed with DI water and ethanol and dried at 60 ◦ C. The dried catalyst was then placed in a tube furnace and charged with O18 -labeled oxygen and calcined at 400 ◦ C for 4 h. Crystallographic analysis for the tested were performed by means of XRD measurements in 2 mode on a Bruker AXS D8 diffractometer with CuK␣ (␭ = 0.154056 Å) radiation at 40 kV and 20 mA. XPS was performed on a Thermo Escalab 250 spectrometer. The binding energy was calibrated using C1 s (284.6 eV) as a reference. The as-synthesized CuO-CeO2 morphology was also studied by SEM (JEOL JSM 6700F field emission), TEM and HR-TEM (JEOL JEM-2100F). H2 -TPR was carried out in a quartz fixed-bed reactor

CuO-CeO2 composites were obtained after calcining the Ce-Cu binary precursors synthesized by the hydrothermal process. The morphology and microstructure of as-synthesized CuO-CeO2 were investigated by SEM, TEM and HR-TEM analysis. The low magnification SEM image in Fig. 1a reveals that the samples consists of spherical particles of 200–300 nm in diameter. The porous nature of the as-synthesis CuO-CeO2 is shown in Fig. 1b (high magnification SEM). The spent CuO-CeO2 catalyst (Fig. 1c and d) is also sphere-like, indicating that the morphology of the fresh catalyst is not destroyed after the catalytic reaction. Fig. 2 displays the TEM and HR-TEM images of the as-synthesized and spent CuOCeO2 nanospheres. The TEM images (Fig. 2a) confirms our above deduction that, the nanospheres consist of nanoparticles. These small nanoparticles consisting of the nanospheres have a diameter of ca. 10 nm. HR-TEM image shown in Fig. 2b of the CuO-CeO2 nanospheres give lattice spacing of 0.156 and 0.234 nm, corresponding to (111) planes of cubic CeO2 phase and (111) planes of monoclinic CuO phase, respectively [24,25]. The HR-TEM analysis reveals that the nanospheres are mainly composed of mixed crystallites, preferentially exposing the disordered crystal phases of CuO and CeO2 crystals. The TEM analysis of the spent catalyst (Fig. 2c) confirms the sphere-like morphology of the spent CuOCeO2 catalyst as revealed by the SEM analysis. The HR-TEM image in Fig. 2d also shows the presence of Cu (111) with lattice fringe of 0.206 nm, and CeO2 (111) with lattice fringe of 0.156 nm. Surface area analysis (BET) is performed to further investigate the porous structure and surface area of the as-synthesized CuO-CeO2 nanospheres. The specific surface area estimated from the BET

Please cite this article in press as: P.N. Amaniampong, et al., Porous structured CuO-CeO2 nanospheres for the direct oxidation of cellobiose and glucose to gluconic acid, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2017.01.009

G Model CATTOD-10547; No. of Pages 11

ARTICLE IN PRESS P.N. Amaniampong et al. / Catalysis Today xxx (2016) xxx–xxx

3

Fig. 1. nanospheres (a) low magnification (b) high magnification SEM images of the as-synthesized CuO-CeO2 (c) low magnification (b) high magnification SEM of the spent CuO-CeO2 catalyst.

method is 51 m2 g−1 , which is evidently superior to that of pure CuO (4 m2 g−1 ) and CeO2 (7 m2 g−1 ) nanospheres prepared using the same hydrothermal method. The improved surface area for the as-synthesized CuO-CeO2 may result from their highly porous structure. Analyses on the XRD pattern are performed on the fresh, spent and oxidative regenerated CuO-CeO2 catalyst. The XRD pattern of the CuO-CeO2 fresh composites is shown in Fig. 3a. All diffraction peaks can be assigned to the cubic phase of cerium (IV) oxide, dominated by CeO2 (111) surface with lattice constant ␣ = 5.4138 Å coexisting with the Tenorite monoclinic phase of copper (II) oxide dominated by CuO (111) and CuO (−111), with typical reflections centered at 35.6◦ , 38.7◦ , 61.5◦ and 66.2◦ . Due to the high calcination temperature, it is possible that a small amount of CuO has been incorporated into the CeO2 lattice. The XRD pattern of the spent CuO-CeO2 (Fig. 3b) reveals diffraction peaks at 2 = 43.5, 50.6 and 74.1◦ , attributed to the (111), (200), and (220) facets of pure copper (Cu) face-centered cubic structure (fcc), respectively. While all other indexed peaks are typical diffractions of CeO2 . The XRD pattern analysis of the oxidative regenerated CuO18 -CeO2 with isotope labeled oxygen (18 O) catalyst (Fig. 3c) shows identical diffraction peaks as the as-synthesized nanospheres composites. The XPS spectra of Ce 3d and Cu 2p binding energies of CuO-CeO2 nanospheres were shown in Fig. 4. The peaks centered at 897.6 and 916.0 eV were ascribed to Ce4+ and those centered at 881.5 and 900.1 eV to Ce3+ (Fig. 4b), suggesting that Ce4+ /Ce3+ redox pair existed in the in the as-synthesized CuO-CeO2 nanospheres [26]. For the binding energy of Cu 2p (Fig. 4c) the presence of the shakeup peak at about 941.6 eV and the lower Cu 2p3/2 binding energy appearing at 932.9 eV were two major XPS characteristics of CuO and Cu2 O [27–29], clearly indicating the presence of Cu2+ /Cu+ redox pair in the as-synthesized CuO-CeO2 nanospheres composite. The

O 1s spectra of the CuO-CeO2 catalyst were shown in Fig. 4d, the main peak centered at a binding energy around 528.9-529.8 eV was characteristic of the lattice oxygen of CeO2 and CuO phases, while the broad shoulder at around 531.5 eV may be assigned to the defect oxide or absorbed oxygen with low coordination [27,30]. The lattice structure and lattice defects of the CuO-CeO2 nanospheres were further investigated by Raman spectroscopy. The UV-Raman spectra of the CuO-CeO2 nanospheres (Fig. 5a) revealed a broad band with relatively high intensity at about 453 cm−1 , which was assigned to the F2g Raman vibration mode of the cubic fluorite-structure of CeO2 in CuO-CeO2 [31,32]. Whereas the peak at 301 cm−1 can be assigned to the Bg modes of CuO [33]. The bands appearing at about 588 and 1124 cm−1 can be linked to oxygen vacancies in the CeO2 and CuO lattices and the presence of CeO2 defects. The presence of oxygen vacancies was crucial to the production and stability of Cu+ species. Noteworthy, the position of the strong F2g band of the as-synthesized CuO-CeO2 nanospheres deviated from that of pure CeO2 (462 cm−1 ) [34], suggesting that the copper ions incorporated in CeO2 lattice and resulted in the formation of the CuO-CeO2 nanospheres composites. The incorporated copper ions induced the lattice distortion of CeO2 , which influenced the polarizability of the symmetrical stretching mode of the [Ce-O8 ] vibrational unit and led to the observed deviation [35,36]. The UV-Raman spectra of the spent CuO-CeO2 catalyst (Fig. 5b) revealed a broad band with relatively high intensity at about 453 cm−1 , similar to the observations made for the Raman results of the as-synthesized CuOCeO2 fresh nanospheres. The new peak at 600 cm−1 can be liked to oxygen vacancies in the CuOCeO2 lattices and the presence of Cu clusters on CeO2 surfaces [37]. Moreover, there was no Raman signature for the presence of CuO (Raman peaks at 288–298, 330–345 and 621–632 cm−1 )[38,39]. Furthermore, the small peak at 212 cm−1 can be assigned to Cu2 O [40],

Please cite this article in press as: P.N. Amaniampong, et al., Porous structured CuO-CeO2 nanospheres for the direct oxidation of cellobiose and glucose to gluconic acid, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2017.01.009

G Model CATTOD-10547; No. of Pages 11 4

ARTICLE IN PRESS P.N. Amaniampong et al. / Catalysis Today xxx (2016) xxx–xxx

Fig. 2. (a) TEM image; (b) high-resolution TEM of the as-synthesized CuO-CeO2 nanospheres (c) TEM image (b) high-resolution TEM of the spent CuO-CeO2 nanospheres.

suggesting the reduction of CuO in the CuO-CeO2 fresh catalyst after the reaction. This was consistent with our XRD, HR-TEM and XPS results for the formation of CuO-CeO2 nanospheres composites, and also further revealed that the lattice oxygen in CuO, in the CuO-CeO2 catalysts were consumed during the oxidation reaction. The reducibility for the CuO-CeO2 nanospheres was further studied by H2 -TPR. It should be clarified that pure CeO2 showed two well resolved reduction peaks (not shown for brevity). One centered at 510 ◦ C, attributed to the reduction of the surface capping oxygen of ceria, and the other at 790 ◦ C was ascribed to the reduction of bulk oxygen in ceria [41]. The as-synthesized CuOCeO2 nanospheres exhibited three major peaks in the temperature at 141.8 (␣), 189.6 (␤), and 214.6 ◦ C (␥), as shown in Fig. 6. The ␣ peak was attributed to the reduction of large CuO particles interacting weakly with CeO2 , while the ␤ peak was attributed to the reduction of pure bulk CuO phase, and the ␥ peak was attributed to the reduction of highly dispersed CuOx species on the surface of CeO2 , including isolated Cu2+ ions that weakly interacted with CeO2 [42,43]. The reduction of pure bulk CuO has been reported occur at temperature above 300 ◦ C in the literature [44]. The significantly low reduction temperatures as compared to the individual bulk catalysts, and high H2 consumption (2814 ␮mol/g) estimated from the TPR-analysis, suggested the strong Cu-Ce interactions in the as-synthesized CuO-CeO2 nanospheres. From these reduction patterns obtained in this work, it can be concluded that a cooperative effect between CeO2 and CuO resulted in lower reduction temperature of CuO.

3.2. Catalytic performance The porous CuO-CeO2 nanospheres were tested for cellobiose and glucose oxidation in the absence of molecular oxygen. It is worth noting that, other metal oxides were also explored as catalyst for the conversion of cellobiose to gluconic acid, under the same reaction conditions and in the absence of molecular oxygen. The reactor was purged with high purity nitrogen several times prior to the reaction. Fig. 7 showed the distribution of reaction products over the metal oxides investigated. Reactions performed over Fe2 O3 and Fe3 O4 revealed similar reaction product distributions with cellobiose conversions of 24.1% and 22.2%, respectively. There was no formation of gluconic acid, and the main products observed were glucose and fructose with yields of 20.2 and 3.9%, respectively for Fe2 O3 catalysts, and 18.2 and 4.0%, respectively, for Fe3 O4 catalyst, suggesting in the poor oxidation activity of Fe2 O3 and Fe3 O4 at the reaction temperatures used in this study [45]. Similar observations were made for reactions performed over SiO2 , with glucose and fructose yields of 30.1 and 17.5%, though with slightly higher conversion (47.6%). For reactions performed over CeO2 a gluconic acid yield of 2.7% was observed, along with glucose (17.3%) and fructose (2.7%) and a cellobiose conversion of 22.2%. For NiO catalyst, a conversion of 16.7% was achieved, with a mixture of glucose, gluconic acid and formic acid as the reaction products, with yields of 2.1, 7.8 and 6.7%, respectively, implying the ability of the NiO catalyst for C C bond cleavage and partial oxidation reactions [46]. Prior to the catalyst investigations using the CuO-CeO2 nanospheres, CuO-CeO2 catalyst loading was investigated. Catalytic

Please cite this article in press as: P.N. Amaniampong, et al., Porous structured CuO-CeO2 nanospheres for the direct oxidation of cellobiose and glucose to gluconic acid, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2017.01.009

G Model CATTOD-10547; No. of Pages 11

ARTICLE IN PRESS P.N. Amaniampong et al. / Catalysis Today xxx (2016) xxx–xxx

5

Table 1 Time course for cellobiose conversion and gluconic acid yield. Reaction conditions: 0.050 g of catalyst, 0.205 g of cellobiose, 15 mL of H2 O, 160 ◦ C reaction temperature. Reaction time [min]

Catalystc

Conversion [%]

Gluconic acid yield [%]

30 60 120 180 180 180 180 240 240a 240b

CuO-CeO2 fresh CuO-CeO2 fresh CuO-CeO2 fresh CuO-CeO2 fresh CuO-CeO2 spent CuO18 -CeO2 reoxidized CuO spent CuO-CeO2 fresh CuO-CeO2 fresh CuO-CeO2 fresh

34 43 67 68 68 69 54 73 70 100

n.d 20 ± 0.2 42 ± 1.0 51 ± 0.8 8 ± 0.2 48 ± 1.8 4 ± 0.5 50 ± 1.2 51 ± 2.5 14 ± 0.6

a

Reactor purged with high purity nitrogen. Reaction temperature = 180 ◦ C. The fresh catalyst refers to the as-synthesized catalysts, reoxidized catalyst refers to the catalyst which after one reaction cycle, is reoxidized for 240 min under isotope labeled oxygen (18 O). Spent catalyst refers to the catalyst which was recovered after undergoing one cycle of reaction. b c

acid formed. Adding CuO-CeO2 catalyst increased the cellobiose conversion from 5.7% observed for the blank reaction to 68.2% with a gluconic acid yield of 50. 7%. Kinetic study of the reaction with CuO-CeO2 fresh catalyst showed that, the conversion of

Fig. 3. (a) XRD pattern of fresh CuO-CeO2 nanospheres (b) XRD pattern of spent CuO-CeO2 nanospheres (c) XRD pattern of reoxidized CuO-CeO2 with isotope labeled oxygen (18 O).

performance of the porous structured CuO-CeO2 nanospheres was evaluated in the oxidation of cellobiose to gluconic acid in terms of cellobiose conversion (%) and gluconic acid yield (%) by varying the catalyst loading. The optimum catalyst loading was observed to be 25 wt.% of CuO-CeO2 related to the substrate (Fig. 8). When lower amounts of catalyst were used in the reaction, lower conversion and yields of gluconic acid were obtained, due to the lower number of oxygen in the catalyst than the number of substrate molecules. However, with the gradual increase in the relative number of CuO oxygen (with increasing CuO-CeO2 amount, keeping the amount of cellobiose constant), the conversion and gluconic acid yield increase. At a catalyst loading of above 25 wt.%, oxidative byproducts (glycolic acid) was observed, leading to a decrease in the yield of gluconic acid. This is due to the over-oxidation of gluconic acid that once formed, competes with the substrate for oxidizing species available in the reaction media. After the catalyst loading optimization, the CuO-CeO2 nanospheres were employed directly for the catalytic oxidation of cellobiose to gluconic acid without any in situ functionalization or post-modification. The conversions of cellobiose at different reaction conditions are shown in Table 1. Noteworthy, a blank reaction was performed at 160 ◦ C for 180 min in the absence of any catalyst, a cellobiose conversion of 5.7% and glucose yield of 5.7% (100% selectivity) was observed, with no traces of gluconic

Fig. 4. XPS spectra of the as-synthesized CuO-CeO2 nanospheres (a) Survey scan (b) Ce 3d (c) Cu 2p (d) O 1s.

Please cite this article in press as: P.N. Amaniampong, et al., Porous structured CuO-CeO2 nanospheres for the direct oxidation of cellobiose and glucose to gluconic acid, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2017.01.009

G Model CATTOD-10547; No. of Pages 11

ARTICLE IN PRESS P.N. Amaniampong et al. / Catalysis Today xxx (2016) xxx–xxx

6

Fig. 4. (Continued) Fig. 5. UV-Raman spectra of (a) CuO-CeO2 fresh catalyst (b) CuO-CeO2 spent catalyst.

cellobiose increased sharply at the initial stage, reaching 66.7% in 120 min, followed by a gradual increase slowly and reaching 73.3% after 240 min. Gluconic acid yield was only observed after 60 min (∼20%) and increased steadily to ∼51% at 180 min. The spent CuO-CeO2 catalyst without regeneration only afforded the formation of glycolic acid and formic acid as major products, with small amounts of gluconic acid yield (8.1%). A dramatic increase in the formation of gluconic acid (48.3% yield) was observed after the oxidative regeneration of the catalyst, with a conversion of 69%, comparable to the results obtained over the fresh catalyst. An increase in the conversion of cellobiose with no significant increase in the yield of gluconic acid was also observed at 240 min. The conversion of cellobiose and yield of gluconic acid did not change significantly, irrespective of the reactor being purged by nitrogen or not, suggesting that dissolved oxygen did not play any role in the reaction and that CuO-CeO2 was the source of oxygen supply in the oxidation reaction. At reaction temperatures higher than 160 ◦ C, complete conversion of cellobiose occurred, with lower yield of gluconic acid (14.3%) and significant amounts of smaller acids (glycolic acid yield of 65.1%), suggesting the cleavage of C C bonds. Compared to reactions performed at 160 ◦ C, the formed gluconic acid at 180 ◦ C suffers from oxidative decomposition and catalytic C C cleavage due to the presence of the reduced

Fig. 6. H2-TPR profile of the as-synthesized CuO-CeO2 nanospheres.

Please cite this article in press as: P.N. Amaniampong, et al., Porous structured CuO-CeO2 nanospheres for the direct oxidation of cellobiose and glucose to gluconic acid, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2017.01.009

G Model CATTOD-10547; No. of Pages 11

ARTICLE IN PRESS P.N. Amaniampong et al. / Catalysis Today xxx (2016) xxx–xxx

Fig. 7. Effect of various metal oxides on the yield of gluconic acid. Reaction conditions: 0.050 g of catalyst, 0.205 g of cellobiose, 15 mL of H2 O, 160 ◦ C reaction temperature.

CuO-CeO2 catalyst at elevated temperatures. This explains the lower yield of gluconic acid at temperatures higher than 160 ◦ C 3.3. Oxygen and hydrogen labeling The CuO18 -CeO2 catalyzed cellobiose to gluconic acid oxidation reaction was performed in water, with the reactor being purged with high purity nitrogen several times to ensure complete removal of any oxygen in the reactor. Product analysis using mass spectroscopy (Fig. 9) revealed the insertion of O18 lattice oxygen from the CuO in CuO18 -CeO2 nanospheres composite in gluconic acid, suggesting that the lattice oxygen in CuO and not CeO2 in the CuO-CeO2 were directly involved in the oxidation reaction. This further confirmed the theoretical prediction with CuO catalyst for the direct oxidation of glucose [6]. The behavior of cellobiose oxidation observed in Scheme 1 suggested that the reaction was carried out following the Mars-van Krevelen mechanism [47]. Surface lattice oxygen atoms from the catalyst were consumed by the C6 H12 O6 (reaction intermediate) molecules, and oxygen vacancies were created. These vacancies were then filled by the oxygen atoms that diffused from the bulk of the catalyst to the surface. Oxygen labeling results suggested that, CeO2 was added to CuO to increase the surface areas, thermal stability, to improve the oxygen mobility within the catalyst lattice [48] and also enhance the redox behavior of copper ions by showing more stable activity compared to pure CeO2 and CuO catalysts. The promotional effect of ceria on copper oxide in the total oxidation of CO has also been reported by Park and Ledford [49].

7

Fig. 8. Influence of catalyst loading on the oxidation of cellobiose to gluconic acid. Reaction conditions: 0.205 g of cellobiose, 15 mL of H2 O, 160 ◦ C reaction temperature, 180 min reaction time.

To obtain more insights into the role of solvent in the reaction, deuterium oxide (D2 O) was used as solvent instead of H2 O. Glucose (the reaction intermediate) was used as the substrate for the oxidative conversion over fresh CuO-CeO2 nanospheres for easy elucidation of the reaction mechanism. Fig. 10 showed the ESI–MS spectrum of the reaction product. The major reaction product was observed to be gluconic acid with “H”, without the insertion of deuterium (D) into the reaction product in the presence of D2 O solvent. An isotopic effect was not observed when the reaction was performed in D2 O, suggesting that H-D exchange processes are not involved in the oxidative conversion of glucose over CuOCeO2 catalyst. The results excluded the oxidative dehydrogenation mechanism for the formation of gluconic acid, where glucose is converted into the hydrate form in aqueous solution and adsorbed on the catalyst surface, the hydrate is dehydrogenated and eventually [21]. To gain further insight into the oxidation reaction mechanism, isotopically labeled glucose was reacted with the CuO-CeO2 catalyst and investigated using 13 C nuclear magnetic resonance (NMR). Before reaction, 13-C NMR analysis the initial d-(1-13-C) glucose solution generated resonances at ␦ = 95.88 and 92.06 ppm, which corresponded to the 13 C at C1 position of the ␤-pyranose and ␣pyranose configurations of glucose, respectively. Their locations were consistent with NMR features in previous reports for d-(1-13 C) glucose [50,51]. After the oxidative reaction, a new additional set of peaks at ␦ = 63.85 and 62.63 ppm and another at ␦ = 180.50 ppm were observed (Fig. 11b). Previous reports have demonstrated that these new peaks, ␦ = 180.50 ppm corresponded to the presence of C1 position of gluconic acid [52]. This confirms the transformation of the aldehyde group of glucose to a carboxylic group. The

Scheme 1. Simplified reaction scheme for the direct oxidation of cellobiose catalyzed by CuO-CeO2 nanospheres.

Please cite this article in press as: P.N. Amaniampong, et al., Porous structured CuO-CeO2 nanospheres for the direct oxidation of cellobiose and glucose to gluconic acid, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2017.01.009

G Model CATTOD-10547; No. of Pages 11 8

ARTICLE IN PRESS P.N. Amaniampong et al. / Catalysis Today xxx (2016) xxx–xxx

Fig. 9. ESI-MS Spectra of the reaction performed with glucose. The gluconic acid (with 18 O oxygen) peak is highlighted in red.(For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 10. ESI-MS Spectra of the reaction performed in D2 O solvent. The gluconic acid (with 16 O oxygen) peak is highlighted in red.(For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Please cite this article in press as: P.N. Amaniampong, et al., Porous structured CuO-CeO2 nanospheres for the direct oxidation of cellobiose and glucose to gluconic acid, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2017.01.009

G Model CATTOD-10547; No. of Pages 11

ARTICLE IN PRESS P.N. Amaniampong et al. / Catalysis Today xxx (2016) xxx–xxx

9

Fig. 11. 13 C NMR spectra of d-(1-13 C) glucose solution (a) before reaction (b) after reaction. Peaks corresponding to d-(1-13 C) glucose are highlighted by a red dotted rectangle. Peaks corresponding to gluconic acid are highlighted by a black dashed rectangle. Peaks corresponding to isomerization product of d-(1-13 C) glucose (d-(1-13 C) fructose) are highlighted by a blue dotted line. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Please cite this article in press as: P.N. Amaniampong, et al., Porous structured CuO-CeO2 nanospheres for the direct oxidation of cellobiose and glucose to gluconic acid, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2017.01.009

G Model CATTOD-10547; No. of Pages 11 10

ARTICLE IN PRESS P.N. Amaniampong et al. / Catalysis Today xxx (2016) xxx–xxx

peaks at ␦ = 63.85 and 62.63 ppm corresponded to the presence of ␣-furanose and ␤-pyranose conformations of d-(1-13 C) fructose [50]. 4. Discussion The results of this study show that the total oxidation of cellobiose over the CuO-CeO2 catalyst was carried out by a redox, i.e., Mars-van Krevelen mechanism. Our characterization results clearly revealed that, only the lattice oxygen of the CuO in the CuO-CeO2 nanospheres composite was consumed in the oxidation reaction and not the lattice oxygen of CeO2 . The XRD patterns of the asprepared and spent catalysts were in perfect agreement with our TEM and HR-TEM analysis, suggesting the reduction of CuO to Cu in CuO-CeO2 catalyst during the direct oxidation of cellobiose to gluconic acid. It is generally accepted that the activity of a catalyst is closely related to its structure and morphology [11], in which porous structure and high specific surface area was particularly effective to enhanced its activity. This phenomenon was also observed for the direct oxidation of cellobiose over CuO-CeO2 catalyst. Although characterization results prove that CeO2 was not directly involved in the oxidation reaction, the addition of ceria to CuO exhibit an enhanced surface area, and the reduction temperature of CuO was lower due to the presence of the CeO2 support. This suggest that, oxygen is bound loosely in CuO in the presence of the CeO2 support and a loosely bound oxygen is more active for cleaving the C H bond [53] of glucose, which is a crucial step for the formation of gluconic acid. The enhanced formation of gluconic acid (∼51%) over the as-synthesized catalyst at 180 min, compared to the formation of gluconic acid (∼8%) over the spent CuO-CeO2 catalyst, further explain that it was indeed the surface lattice oxygen of the CuO in the CuO-CeO2 catalyst that oxidizes the reactant to form gluconic acid and not the surface lattice oxygen of the CeO2 in CuO-CeO2 catalyst or even both. However, upon the oxidative (18 O) regeneration of the spent catalyst, the catalyst regained its selectivity towards gluconic acid. The conversion of cellobiose and gluconic acid yield was comparable to those of the fresh catalysts as shown in Table 1. The XRD analysis of the oxidative (18 O) regenerative catalyst showed that the reduced CuO in the CuO-CeO2 , are reoxidized. Indeed, the formation of gluconic acid over supported noble metals has been widely been reported to occur via the classical oxidative dehydrogenation mechanism [21,54]. In which open chain glucose converts in aqueous solution into the hydrate form, followed by adsorption onto the catalyst surface, dehydrogenation of the hydrate and eventual desorption. This reaction mechanism show the participatory role of water (solvent) in assisting the protonation of the adsorbed gluconate to form gluconic acid. However, this reaction mechanism was not observed in this study. The hydrogen labeled experiment, where D2 O was used as solvent instead of H2 O did not reveal the insertion of deuterium in the final product (gluconic acid). The formation of gluconic acid as revealed by the 13 C NMR analysis and the presence of fructose as shown in the NMR results, confirmed the direct oxidation of the aldehyde functional group (R-CHO) of glucose to gluconic acid and isomerization of glucose to fructose, by the surface lattice oxygen of CuO in CuO-CeO2 catalysts and the Lewis acid sites in CuO-CeO2 , respectively. 5. Conclusion In this work, CuO-CeO2 nanospheres were synthesized by a hydrothermal treatment and employed in the direct oxidation of cellobiose to gluconic acid. The presence of CeO2 in the CuO-CeO2 mixed-oxide composite results in a highly porous structure with enhanced surface area and also enhances the redox behavior of the

copper ions. The oxidation of cellobiose to gluconic acid on CuOCeO2 nanospheres proceeds by the hydrolysis of cellobiose via the Lewis acid sites of the catalyst, perpendicular end-on adsorption of glucose (cellobiose hydrolysis product) on surface lattice oxygen atoms of CuO-CeO2 catalyst, abstraction of H-atoms from the formyl group, formation of gluconate and the hydrogenation of the gluconate to gluconic acid. 18 O-oxgen isotope labeled experiments coupled with XRD, HR-TEM, XPS, H2 -TPR and UV-Raman spectra analysis revealed that the lattice oxygen of the CuO framework, in the CuO-CeO2 nanospheres composites, and not the lattice oxygen from CeO2 , was consumed during the oxidation reaction to form gluconic acid. ESI-MS and 13 C NMR analyses revealed that, indeed surface lattice oxygen was incorporated into glucose to form gluconic acid. The Cu+ species finely dispersed on the surface of CeO2 , oxygen vacancies and lattice oxygen are advantageous for promoting the catalytic performance of the CuO-CeO2 catalysts for the selective oxidation of cellobiose to gluconic acid. Conflicts of interest The authors declare no conflicts of interest. Acknowledgements We acknowledge the financial support from the National Research Foundation (NRF), Prime Minister’s Office, Singapore under its Campus for Research Excellence and Technological Enterprise (CREATE) program and AcRF Tier 1 grant (RG129/14), Ministry of Education, Singapore. References [1] G.J.d.A.A. Soler-Illia, C. Sanchez, B. Lebeau, J. Patarin, Chem. Rev. 102 (2002) 4093–4138. [2] X. Xie, Y. Li, Z.-Q. Liu, M. Haruta, W. Shen, Nature 458 (2009) 746–749. [3] J.Y. Li, S. Xiong, B. Xi, X.G. Li, Y.T. Qian, Cryst. Growth Des. 9 (2009) 4108–4115. [4] M. Vaseem, A. Umar, Y.B. Hahn, D.H. Kim, K.S. Lee, J.S. Jang, J.S. Lee, Catal. Commun. 10 (2008) 11–16. [5] X. Zhang, G. Wang, X. Liu, J. Wu, M. Li, J. Gu, H. Liu, B. Fang, J. Phys. Chem. C 112 (2008) 16845–16849. [6] P.N. Amaniampong, Q.T. Trinh, B. Wang, A. Borgna, Y. Yang, S.H. Mushrif, Angew. Chem. 127 (2015) 9056–9061. [7] Q. Zhang, K. Zhang, D. Xu, G. Yang, H. Huang, F. Nie, C. Liu, S. Yang, Prog. Mater. Sci. 60 (2014) 208–337. [8] H. Rajesh, U.S. Ozkan, Ind. Eng. Chem. Res. 32 (1993) 1622–1630. [9] P.-O. Larsson, A. Andersson, J. Catal. 179 (1998) 72–89. [10] P.-O. Larsson, A. Andersson, Appl. Catal. B: Environ. 24 (2000) 175–192. [11] C.-H. Wang, S.-S. Lin, C.-L. Chen, H.-S. Weng, Chemosphere 64 (2006) 503–509. [12] P.W. Park, J.S. Ledford, Appl. Catal. B: Environ. 15 (1998) 221–231. [13] L. Kundakovic, M. Flytzani-Stephanopoulos, Appl. Catal. A: Gen. 171 (1998) 13–29. [14] D.C. Sayle, S.A. Maicaneanu, G.W. Watson, J. Am. Chem. Soc. 124 (2002) 11429–11439. [15] M. Luo, J. Chen, L. Chen, J. Lu, Z. Feng, C. Li, Chem. Mater. 13 (2001) 197–202. [16] A. Bielanski, J. Haber, Oxygen in Catalysis, CRC Press, 1990. [17] Q.T. Trinh, B.K. Chethana, S.H. Mushrif, J. Phys. Chem. C 119 (2015) 17137–17145. [18] E. Finocchio, G. Busca, V. Lorenzelli, R.J. Willey, J. Chem. Soc. Faraday Trans. 90 (1994) 3347–3356. [19] H. Wittcoff, Chem. Tech. 20 (1990) 48–53. [20] C. Della Pina, E. Falletta, L. Prati, M. Rossi, Chem. Soc. Rev. 37 (2008) 2077–2095. [21] Y. Önal, S. Schimpf, P. Claus, J. Catal. 223 (2004) 122–133. [22] J. Qin, J. Lu, M. Cao, C. Hu, Nanoscale 2 (2010) 2739–2743. [23] B.J. Johnson, E.A. Betterton, D. Craig, J. Atmos. Chem. 24 (1996) 113–119. [24] S. Jacobsen, U. Helmersson, R. Erlandsson, B. Skårman, L. Wallenberg, Surf. Sci. 429 (1999) 22–33. [25] L.-Y. Kuo, P. Shen, Mater. Sci. Eng.: A 277 (2000) 258–265. [26] M.-F. Luo, Y.-P. Song, J.-Q. Lu, X.-Y. Wang, Z.-Y. Pu, J. Phys. Chem. C 111 (2007) 12686–12692. [27] G. Avgouropoulos, T. Ioannides, Appl. Catal. B: Environ. 67 (2006) 1–11. [28] W. Liu, M. Flytzanistephanopoulos, J. Catal. 153 (1995) 304–316. [29] G. Avgouropoulos, T. Ioannides, Appl. Catal. A: Gen. 244 (2003) 155–167. [30] J. Li, P. Zhu, S. Zuo, Q. Huang, R. Zhou, Appl. Catal. A: Gen. 381 (2010) 261–266. [31] J. Fan, X. Wu, X. Wu, Q. Liang, R. Ran, D. Weng, Appl. Catal. B: Environ. 81 (2008) 38–48.

Please cite this article in press as: P.N. Amaniampong, et al., Porous structured CuO-CeO2 nanospheres for the direct oxidation of cellobiose and glucose to gluconic acid, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2017.01.009

G Model CATTOD-10547; No. of Pages 11

ARTICLE IN PRESS P.N. Amaniampong et al. / Catalysis Today xxx (2016) xxx–xxx

[32] B.M. Reddy, A. Khan, Y. Yamada, T. Kobayashi, S. Loridant, J.-C. Volta, J. Phys. Chem. B 107 (2003) 5162–5167. [33] P. Chand, A. Gaur, A. Kumar, Acta Metall. Sin. (English Lett.) 27 (2014) 306–312. [34] W. Shan, W. Shen, C. Li, Chem. Mater. 15 (2003) 4761–4767. [35] J. Li, Y. Han, Y. Zhu, R. Zhou, Appl. Catal. B: Environ. 108 (2011) 72–80. [36] Y. She, Q. Zheng, L. Li, Y. Zhan, C. Chen, Y. Zheng, X. Lin, Int. J. Hydrogen Energy 34 (2009) 8929–8936. [37] X. Wang, J.A. Rodriguez, J.C. Hanson, D. Gamarra, A. Martínez-Arias, M. Fernández-García, J. Phys. Chem. B 109 (2005) 19595–19603. [38] J. Chrzanowski, J. Irwin, Solid State Commun. 70 (1989) 11–14. [39] J. Xu, W. Ji, Z. Shen, W. Li, S. Tang, X. Ye, D. Jia, X. Xin, J. Raman Spectrosc. 30 (1999) 413–415. [40] D. Powell, A. Compaan, J. Macdonald, R. Forman, Phys. Rev. B 12 (1975) 20. [41] S. Damyanova, B. Pawelec, K. Arishtirova, M.M. Huerta, J. Fierro, Appl. Catal. A: Gen. 337 (2008) 86–96. [42] P. Zhu, J. Li, S. Zuo, R. Zhou, Appl. Surf. Sci. 255 (2008) 2903–2909. [43] X.-C. Zheng, S.-H. Wu, S.-P. Wang, S.-R. Wang, S.-M. Zhang, W.-P. Huang, Appl. Catal. A: Gen. 283 (2005) 217–223.

11

[44] J. Xiaoyuan, L. Guanglie, Z. Renxian, M. Jianxin, C. Yu, Z. Xiaoming, Appl. Surf. Sci. 173 (2001) 208–220. [45] Z. Zhong, J. Highfield, M. Lin, J. Teo, Y.-f. Han, Langmuir 24 (2008) 8576–8582. [46] E. Ruckenstein, Y.H. Hu, Appl. Catal. A: Gen. 183 (1999) 85–92. [47] P. Mars, D.W. van Krevelen, Chem. Eng. Sci. 3 (Suppl. 1) (1954) 41–59. ´ J. Batista, A. Pintar, Appl. Catal. A: Gen. 347 (2008) 23–33. [48] P. Djinovic, [49] P.W. Park, J.S. Ledford, Catal. Lett. 50 (1998) 41–48. [50] W.R. Gunther, Y. Wang, Y. Ji, V.K. Michaelis, S.T. Hunt, R.G. Griffin, Y. Román-Leshkov, Nat. Commun. 3 (2012) 1109. [51] Z.C. Gernhart, A. Bhalkikar, J.J. Burke, K.O. Sonnenfeld, C.M. Marin, R. Zbasnik, C.L. Cheung, RSC Adv. 5 (2015) 28478–28486. [52] E.L. Ulrich, H. Akutsu, J.F. Doreleijers, Y. Harano, Y.E. Ioannidis, J. Lin, M. Livny, S. Mading, D. Maziuk, Z. Miller, Nucl. Acids Res. 36 (2008) D402–D408. [53] J.J. Varghese, Q.T. Trinh, S.H. Mushrif, Catal. Sci. Technol. 6 (2016) 3984–3996. [54] P.N. Amaniampong, A.Y. Booshehri, X. Jia, Y. Dai, B. Wang, S.H. Mushrif, A. Borgna, Y. Yang, Appl. Catal. A: Gen. 505 (2015) 16–27.

Please cite this article in press as: P.N. Amaniampong, et al., Porous structured CuO-CeO2 nanospheres for the direct oxidation of cellobiose and glucose to gluconic acid, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2017.01.009