3.1. Catalyst characterization
Figure 1b shows the powder XRD pattern for the metal oxide catalysts. The wide angle XRD patterns of the sample of La2O3 showed reflections at 2θ = 15.70º, 27.18º, 28.09º, 39.41º, 48.2º ascribed to La2O3 phase [JCPDS # 05-0602] corresponding to (100), (002), (101), (102), (211) planes, respectively, with lattice constants as a = 3.397 nm and b = 6.129 nm which indicated the presence of hexagonal phase of La2O3 [JCPDS # 05-0602].(37) The diffraction peaks at 2θ = 15.70º and 48.80º can be indexed to the hexagonal phase of La(OH)3 corresponding to (100) and (211) planes with lattice constants as a = 0.652 nm and b = 0.858 nm.(38) The appearance of sharp peaks for the hexagonal phases of La2O3 and LaOH3 denote high crystallinity of La2O3. La2O3 didn’t show typical peak for carbonated form of lanthanum which can be observed at 2θ = 10.2o, it confirmed the presence of La2O3 phase and small amount of La(OH)3 phase in the examined sample.(39) All the other catalysts used for comparison also showed well defined phase. The wide angle XRD pattern for tetragonal phase of BaO showed the reflections at 2θ = 19.6º, 23.7º, 45.1°, [JCPDS#47-1488]. Whereas, reflections at 2θ = 23.7º, 29.1º and 55.8º were ascribed to monoclinic ZrO2 phase [JCPDS#86-1451].(40) The XRD pattern for CaO showed peaks at 18.14º, 28.73º, 34.16º, 47.11º, 50.86º, 54.36º, 62.76º, 64.41º, 72.05º, and 84.87º which confirmed the presence of CaO phase.(41) The cubic phase of CeO2 as confirmed by the presence of the characteristic peaks observed at 2θ = 28.83º,33.2º, 47.9º, 56.7º and 59.4º corresponding to (111), (200), (220), (311) and (222) planes, respectively.(42) The XRD pattern for Al2O3 showed weak diffraction peaks, observed at 2θ = 19.2º, 31.0º, 36.6º, 39.3º, 46º, 61.5º, and 67º, which can be indexed to the reflections at (111), (220), (311), (222), (400), (511), and (440) corresponding to the presence of γ-alumina phase according to JCPDS card: 100425.(43) The detailed morphology, surface topography of the catalysts was studied FE- SEM.
Figure 2 displays the SEM images for various catalysts, in different magnifications viz. (a) CaO, (b) BaO, (c) La2O3, (d) Ce2O3. In case of calcined La2O3, irregular and uneven shaped particles were predominantly present as shown in Fig. 2c. The SEM images clearly indicated the presence of pores of varied sizes in the range 90–450 nm on the surface. The presence of these nano pores aid in providing high surface area for adsorption. The SEM analysis of cerium oxide is shown in Fig. 2d.The morphological investigation revealed the presence of pores with spherical shaped particles distributed on the surface with size 90–420 nm. BaO showed the presence of mixed morphology with spherical and rod shaped particles. Rod shaped particles were present predominantly with length in the range of 1-2.8 µm and the size of the spherical particles in the range 220–450 nm as shown in Fig. 2b.(40) In case of CaO as shown in Fig. 2a, the morphology revealed to contain clusters of flakey particles with particle size, viz. 150–750 nm. The percentage elemental composition was acquired by EDS analysis for all four metal oxide catalysts screened (Supporting information Fig.sS-10, S-11, S-12, S-13).It confirmed the presence of La2O3 phase having composition in mass percentage of the elements as 39.69% O and 60.31% La by weight. The peak indexing of the elements for lanthanum was 4.63 KeV and for oxygen was 0.51 KeV.
High Resolution-Transmission Electron Microscopy (HR-TEM) was utilized for detailed study of the size, shape, morphology and surface distribution of the particles as shown in Fig. 3 Irregular, unordered distributions of the particles were observed at the surface. The agglomeration of the nano particles due to aggregation of the particles by weak forces was observed with average pore size in the range 90–500 nm.(44)Thus, from the HR-TEM characterization we conclude that the agglomerated catalyst possess good porosity due to the presence of pores in the nano range.(45)
Figure 4A presents FT-IR spectra of calcined (La2O3) in the range 400–4000 cm-1 which showed detailed information about the metal and oxygen bond present in the metal oxide catalysts. The FTIR spectrum of La2O3 sample showed a prominent band at 637 cm-1 which was assigned to the stretching vibration of La-O.The intense and sharp absorption band at 3619 cm-1 was ascribed to the -OH stretching of water molecule absorbed from atmosphere on the oxide surface and the band at 1530 cm-1 corresponds to the presence of extending and twisting -OH bending vibration due to the physically adsorbed water molecule on the catalyst surface. This confirms the presence of Brønsted basic sites in the form of La(OH)3 phase.(46–48) Hence, the appearance of the above mentioned bands in the FTIR spectra confirmed the presence of La2O3 and the La(OH)3 phases in accordance with XRD study.(49) BaO showed an intense band at 692 cm-1 correspondingto Ba-O stretching frequency. Whereas, the band at 512 cm-1 was assigned to the Ca-O stretching vibration. The Ce-O stretching frequency was observed at 613 cm-1. In case of CaO and BaO, the absorption band due to the bending vibration of the adsorbed water molecule on the catalyst surface was observed at 1448 cm-1 and intense and sharp absorption due to the -OH stretching of the adsorbed water molecule was observed at 3645 cm-1(sharp) and 3350–3655 cm-1 (broad). The broad absorption band corresponds to the superposition of the hydroxyl stretching bands due to the hydroxyl groups present at the metal oxide surface and the hydrogen bond. But in Ce2O3, the band due to bending vibration for the adsorbed water molecule shifted to 1626 cm-1 and a broad absorption band at 3200–3650 cm-1was observed due to absorbed water molecule at the catalytic surface. This clearly indicates the presence of hydroxyl group on the catalytic surface.
Figure 4B represents the in-situ FT-IR spectra obtained by subtracting adsorbed MeOH on metal oxide and neat metal oxide spectra. The investigation of Brønsted basicity was carried out by performing in-situ methanol adsorption-IR spectroscopy. Earlier, Verneker et al. studied various interactions of FeO(OH)at the surface by MeOH adsorption IR spectroscopy. Their results showed that MeOH interacts with the basic sites on the catalytic surface to form monodentate methoxy species, bidentate methoxy species and molecularly adsorbed species with bands observed at 1115 cm-1, 1092 cm-1, 1064 cm-1.(50–52) In Fig. 4B,various interactions of the adsorbed MeOH with the catalytic surface can be seen. The formation of H-bonded molecularly adsorbed MeOH at the La2O3 surface was observed with a band at 1064 cm-1.Whereas, the bands at 1115 cm-1and 1092 cm-1 indicated the formation of monodentate and bidentate methoxy species/ metal complexed methoxy species. All these bands were observed in La2O3, Ce2O3, CaO and BaO catalysts. The La2O3 catalyst showed intense band at 1063 cm-1 and 1117 cm-1 whereas, CaO and BaO showed weak bands in this region. The presence of these intense bands in La2O3 catalyst clearly indicates the presence of Brønsted basic sites on the catalyst surface, which play an important role in the NAG dehydration.
Table 1 reveals the physico chemical properties of the screened catalysts. The surface areas of the catalysts determined by the Brunauer–Emmett–Teller (BET) measurement using N2 adsorption desorption isotherms, were found to be 98.097 m2g− 1 for calcium oxide, 8.073 m2g− 1 for barium oxide, 12.186 m2g− 1 for lanthanum oxide and 13.294 m2g− 1 for cerium oxide catalyst. Fig.S-8 (supporting info.) represents the adsorption isotherm of all the catalysts. The N2 adsorption- desorption isotherm gives detailed information of surface area, pore volume and pore size of the catalyst calculated using BJH method. The La2O3 catalyst exhibited adsorption isotherm which represents unrestricted multilayer formation due to strong lateral interactions between the adsorbed molecules.(53) The steepness of the isotherm was found to decrease in the order BaO > CaO > La2O3 = Ce2O3. The hysteresis loop in case of CaO showed the presence of relatively uniform and narrow pores whereas, the hysteresis loop for La2O3and Ce2O3 showed the presence of narrow pores with irregular shape and size.(54–55)Table 1shows the BET surface area in terms of (m2g-1), it showed the total surface area of 12.186 m2/g, pore volume of 0.0413 cc/g and pore diameter of 0.1357 nm.(56–57) whereas, Ce2O3 showed total surface area of 13.294 m2g-1, pore volume of 0.0405 ccg-1 and pore diameter of 0.1219 nm. Also, BaO and CaO exhibited total surface area of 8.073 m2/g and 98.097 m2/g, pore volume of 0.0132 cc/g and 0.1569 cc/g and pore diameter of 0.6582 nm and 0.6398 nm respectively.
Table 1 showed the total CO2 desorbed in terms of mmol/gm as well as the temperature wise distribution of the basic sites at the catalytic surface for all the samples under study. It was used to determine the surface basicity of the catalytic samples. The catalytic samples exhibited three CO2 desorption peaks corresponding to the weak, moderate and strong Brønsted basicity. The first peak in the region of 100–200 ºC represents weak Brønsted basicity corresponding to the desorption of physisorbed CO2. Whereas, the peak in the temperature range 200–400 ºC represents moderate basic strength and that in the range 400–700 ºC represents strong basicity. The orders of the total basicity of the screened catalysts were found to be BaO˃ CaO˃ La2O3˃ Ce2O3. The total basicity of La2O3 was 0.3854 mmol g-1, 0.0867 mmol g-1 for Ce2O3, 0.7533 mmol g-1 for CaO and highest 0.99110 mmol g-1 for BaO catalyst. Fig.S9 (supporting info.) represents the TPD plots for various screened catalysts. The La2O3 catalyst was found to give total basicity around 0.3854 mmol g-1 with peak maxima at 335 ºC. Whereas, CaO was found to give basicity of 0.7533 mmol g-1with peak maxima at 420 ºC, Cerium oxide showed basicity of 0.0867 mmol g-1 with peak maxima at 290 ºC and BaO showed highest basicity of 0.9911 mmol g-1.TheCaO and BaO showed maximum desorption in the higher temperature range. Whereas, La2O3 and Ce2O3 catalyst showed maximum desorption in the medium temperature range. This was due to the –OH moiety at the catalytic surface and the presence of co-ordinated O2- species as evidenced by the MeOH-FTIR (presence of Brønsted sites) and XRD analysis (presence of La(OH3) phase). As, La2O3 catalyst showed strong desorption peak in the low temperature range than Ce2O3 catalyst, indicating the presence of weak but dense basic sites at the catalytic surface amongst all the catalysts. Hence, the basic properties of the metal oxide catalyst were evaluated quantitavely by CO2-TPD measurement.
Table 1
Physico-chemical properties of the catalysts
Sr. No. | Catalyst | S-[BET]a (m2g-1) | Pore volumeb (cc g-1) | Pore diameterb (nm) | Distribution of basic sites (mmolg-1) | Total Basicityc (total CO2 desorbed, mmol g-1) |
Temperaturec (°C) (100–200) | Temperaturec (°C) (200–400) | Temperaturec (°C) (400–700) |
1 | CaO | 98.097 | 0.1569 | 0.6398 | - | 0.6780 | 0.0753 | 0.7533 |
2 | BaO | 8.073 | 0.0132 | 0.6582 | 0.0081 | 0.8512 | 0.1318 | 0.9911 |
3 | La2O3 | 12.186 | 0.0413 | 0.1357 | 0.2739 | 0.0940 | 0.0175 | 0.3854 |
4 | Ce2O3 | 13.294 | 0.0405 | 0.1219 | 0.0090 | 0.0068 | 0.0709 | 0.0867 |
a- BET analysis, b- BJH method, c-Determined by CO2 TPD measurement, S-[BET] - BET surface area |
3.2 Catalyst Screening
One pot partial deoxygenation and dehydration of NAG to 3AF and 3A5AF was studied using various metal oxides and the results are presented in Table 2. NAG in absence of any catalyst under the same reaction conditions didn’t show any formation of 3AF and 3A5AF. Interestingly, all the metal oxides showed activities according to acidic and basic nature, they possessed.Al2O3 catalyst gave 11% yield of 3AF and 18% yield towards 3A5AF (Table 2, entry 1). For ZrO2 as a catalyst, NAG yields 12% of 3AF and 6% of 3A5AF (Table 2, entry 2). BaO showed 19% and 5% yield for 3AF and 3A5AF, respectively (Table 2, entry 3). CaO produced 36% yield of 3AF and 19% yield towards 3A5AF (Table 2, entry 4). It can be seen that the basic nature of the catalyst is responsible for formation of the desired product i.e. 3AF as compared to activities shown by acidic metal oxides. The 3AF yield could be enhanced by using rare earth metal oxides which possess appreciable basic characteristics. La2O3 showed almost complete conversion of NAG with 50% yield towards 3AF and 21% yield for 3A5AF (Table 2, entry 5). Another lanthanum series metal oxide Ce2O3 gave 32% 3AF yield and 11% yield for 3A5AF (Table 2, entry 6). It can be inferred from the above results that La2O3 was a very promising catalyst for NAG dehydration and partial deoxygenation producing 3AF product more selectively. La2O3 showed better yield for 3AF as it possesses moderate basicity which are Brønsted in nature along with well distributed three-dimensional interconnected nanopores which can be clearly seen from HR-TEM micrographs. These nanopores present on the surface of La2O3 play a very important role during reaction, allowing improved diffusion of reactants and products. Further using La2O3 as catalyst optimization of various reaction parameters was carried out for achieving the maximum yield for desired 3AF product.
Table 2
Catalyst screening for NAG conversion to 3AF and 3A5AF
Sr. No. | Catalyst | Yield (%) |
3AF | 3A5AF |
1 | Al2O3 | 11 | 18 |
2 | ZrO2 | 12 | 06 |
3 | BaO | 19 | 05 |
4 | CaO | 36 | 19 |
5 | La2O3 | 50 | 21 |
6 | Ce2O3 | 32 | 11 |
Reaction Condition: NAG (1 gm), catalyst (0.2 gm), Dioxane (25 mL), temperature, 180 ºC; reaction time, 3 h |
3.2.1 Effect Of Substrate To Catalyst Ratio
Figure 5 represents the effect of substrate to catalyst ratio on N-acetyl-glucosamine conversion. The catalyst loading was varied in range 5 to 25 wt. % with respect to the substrate (constant at 1 gm) to determine its effect on NAG conversion to 3AF.With increase in the catalyst loading from 5 to 20 wt. % there was an increase in the yield from 20 to 50%, which could be imputed to the availability of nanopores as well as active basic sites at the catalytic surface for better adsorption. As the catalyst loading was further increased to 25 wt. % there was slight decrease in the yield (48%), due to increase in the formation of humins. Thus, 20 wt. % of the catalyst was optimized for desired conversion.
3.2.1 Effect Of Time
Figure 6 shows significant increase in 3AF and 3A5AF production with increase in time 60 to 180 min. 3AF and 3A5AF yields were found to increase linearly with increase in time to maximum upto 50% and 19%, respectively. However, the yield of 3AF and 3A5AF was found to decrease after 180 min due to thermal decomposition of the products and acceleration in the rate of formation of humins as by-product was favoured by elevated temperature and prolonged time.
3.2.3 Effect Of Reaction Temperature
Figure 7 shows the effect of temperature on N-acetyl glucosamine conversion. NAG conversion to 3AF with La2O3 catalyst was studied by examining the reaction at various temperatures in range 120–210 ºC. Figure 8 illustrates that with increase in temperature there was a linear increase in the yield of 3AF and reached to maximum at 180 ºC. With further increase in the temperature to 210 ºC, the product yield decreased as high temperature accelerated the humin formation. Thus, with increase in temperature above 180 ºC, the selectivity to 3AF decreased due to enhancement in the formation of humins. Thus, this decrease in product yield at elevated temperature (above 180 ºC) indicates that high temperature favours humin formation. The increase in temperature causes heat transfer resulting in increase in collisions between the particles thereby increasing the collision frequency. This lowers the energy of activation and enhances the reaction rates. The increasing temperature accelerated the 3AF production. Thus, maximum 50% 3AF yield was observed within 3 h at 180 ºC.
3.2.4 Solvent Screening
Solvents play a vital role in various phenomena such as heat transfer, providing medium for reaction, separation and purification of the products. Figure 8 exhibits the effect of solvent on N-acetyl glucosamine conversion to 3AF. In order to understand the influence of solvent system on NAG conversion to 3AF and 3A5AF, several solvents such as DMA, DMSO, DMF, MIBK and dioxane were screened. DMA, DMSO and DMF were preferred solvents as they were used for dehydration of cellulosic and chitin biomass previously.(58)Whereas, dioxane is an aprotic solvent and is able to solvate various inorganic compounds in it. DMF gave highest product yield of 55%, while DMA, DMSO, MIBK and dioxane gave52%, 48%, 39% and 50% yields of 3AF, respectively. Among these, DMF, DMA and DMSO are toxic and have few health hazards and separation of products is quite tedious as these are high boiling solvents. Also, MIBK showed low selectivity to the product due to formation of large amount of humins. Thus, dioxane was chosen as the best solvent for this conversion.
3.2.5 Effect Of Addition Of Water On Reaction Pathway:
As water being green solvent and chitin biomass being readily soluble in water, the mixtures of above individual solvents with water were screened for the conversion of NAG to 3 AF and 3A5AF. It was observed that the addition of water to the organic phase significantly decreased the product selectivity. Thus, the addition of water to the organic phase showed negative effect on the product yield and selectivity. This could be attributed to the fact that water caused acceleration in the formation of humins thereby decreasing the selectivity of the product formation. The effect of addition of water on NAG conversion was studied by adding 50% water to the solvent system chosen. Figure 9 shows the effect of addition of water to the solvents DMA, DMSO, DMF, MIBK and Dioxane on N-Acetylglucosamine conversion. Thus, water has negative effect on NAG conversion as presence of water enhances the humin formation.
3.2.6 Catalyst Recycle Study
The recycle study of the heterogeneous (La2O3) catalyst was carried out under the optimized conditions and results are as shown in Fig. 10. After the first run, catalyst was filtered and washed with deionised water till the pale-yellow coloured catalyst faded, the catalyst was dried at 110 ºC in an oven and then calcined at 550 ºC for further 3 runs. The procedure was followed for four successive cycles. The N-acetyl glucosamine conversion decreased by marginally by 3% (from 50–47%) due to handling losses. In order to study the changes of the catalyst surface during the reaction, the recovered catalyst was characterised by XRD, FTIR and TGA analysis. The XRD pattern of the used catalyst (Fig. 11) showed that the crystallinity of the catalyst remained intact and there was no deposition of any carbonatious material at the catalytic surface. The FTIR spectrum of the reused catalyst (Fig. 12) was similar with that of the fresh catalyst which confirms the stability of the catalyst. The TGA analysis (Fig.S-14, supporting info.)of the catalysts confirmed that the catalyst was thermally stable. Thus, the La2O3 catalyst showed exquisite recyclability performance up to 4 cycles.