Abstract
The catalytic conversion of biomass into high-value chemicals, such as furan derivatives, from 5-hydroxymethylfurfural (HMF) holds significant importance. In this study, a bifunctional catalyst prepared from phosphomolybdic acid (PMA) and chitosan was developed for the one-pot synthesis of 2,5-diformylfuran (DFF) from fructose. This approach offered the advantage of bypassing the solvent consumption and cost associated with HMF separation and purification for DFF production. The catalytic activity of the prepared catalyst primarily originated from the loaded PMA. Under optimized reaction conditions, the catalyst achieved a DFF yield of approximately 60% after a 6-h reaction at 150 °C or a 1-h reaction at 170 °C. Moreover, the catalyst exhibited good stability and reusability. These results highlight the potential of the bifunctional catalyst for efficient and cost-effective conversion of fructose into DFF, enabling practical applications in the production of valuable chemicals from bioresources.
Download PDF
Full Article
One-Pot Synthesis of 2,5-Diformylfuran from Fructose using a Bifunctional Catalyst Derived from Phosphomolybdic Acid and Chitosan
Davidraj Johnravindar,a Mian Laiq Ur Rehman,a Peixin Wang,a Reeti Kumar,a and Jun Zhao a,b,*
The catalytic conversion of biomass into high-value chemicals, such as furan derivatives, from 5-hydroxymethylfurfural (HMF) holds significant importance. In this study, a bifunctional catalyst prepared from phosphomolybdic acid (PMA) and chitosan was developed for the one-pot synthesis of 2,5-diformylfuran (DFF) from fructose. This approach offered the advantage of bypassing the solvent consumption and cost associated with HMF separation and purification for DFF production. The catalytic activity of the prepared catalyst primarily originated from the loaded PMA. Under optimized reaction conditions, the catalyst achieved a DFF yield of approximately 60% after a 6-h reaction at 150 °C or a 1-h reaction at 170 °C. Moreover, the catalyst exhibited good stability and reusability. These results highlight the potential of the bifunctional catalyst for efficient and cost-effective conversion of fructose into DFF, enabling practical applications in the production of valuable chemicals from bioresources.
DOI: 10.15376/biores.19.2.2948-2966
Keywords: Fructose; 5-Hydroxymethylfurfural; 2,5-Diformylfuran; Phosphomolybdic acid; Chitosan
Contact information: a: Department of Biology, Hong Kong Baptist University, Hong Kong SAR; b: Institute of Advanced Materials, Hong Kong Baptist University, Hong Kong SAR;
* Corresponding author: zhaojun@hkbu.edu.hk
GRAPHICAL ABSTRACT
INTRODUCTION
The utilization of biomass feedstocks to produce high-value chemicals offers a promising solution to mitigate environmental pollution and high carbon emissions associated with petroleum-based industries. The conversion of biomass into value-added chemicals enables the sustainable utilization of renewable resources, reduces dependence on fossil fuels, and contributes to the development of a circular economy (Wang and Tester 2023). Among various biomass conversion reactions, the production of various furan derivatives using biomass-based 5-hydroxymethylfurfural (HMF) as a platform chemical has garnered significant attention for the valorization of biomass resources. The HMF, derived from the dehydration of fructose, possesses active chemical properties that make it an excellent precursor for the synthesis of a wide range of furan-based compounds (Rosatella et al. 2011). These derivatives find extensive applications across various industries, including plastics, pharmaceuticals, and agrochemicals.
One of the most promising derivatives of HMF is the 2,5-diformylfuran (DFF), which can be synthesized through the oxidation of HMF. The DFF serves as a key intermediate for the synthesis of a plethora of compounds, including but not limited to, polymers, fine chemicals, and even potential biofuels (Cajnko et al. 2020; Slak et al. 2022). The synthesis of DFF from HMF predominantly relies on the selective oxidation of the hydroxymethyl group of HMF and various catalysts such as VOPO4.2H2O (Carlini et al. 2005), Mn-Cs (Solanki et al. 2023), Bi2WO6/mpg–C3N4 (Cheng et al. 2021), V/Mn-1 (Zhu et al. 2021), α-MnO2 (Yu et al. 2021), MgO.MnO2.CeO2 (Nocito et al. 2018), NH4·V3O8/Fe3O4 (Lai et al. 2019), P−C−N-5-800 (Zhang et al. 2020), and NC-950 (Ren et al. 2018). These catalysts have been found to display good performance. DFF synthesis via catalytic oxidation of HMF can achieve significant yields, yet the intrinsic reactivity of HMF leads to challenging separation and purification processes. These additional steps contribute to elevated costs, rendering economically unattractive the industrial-scale production of DFF from pure HMF (Trapasso et al. 2022). Moreover, this process is environmentally detrimental due to the extensive use of organic solvents.
The direct production of DFF from ubiquitously available low-cost carbohydrates can overcome these challenges. HMF can be conveniently synthesized through the dehydration of fructose. If HMF could be transformed into downstream products in situ within the same reaction system, bypassing the need for its separation and purification, it would significantly reduce the overall production costs (Šivec et al. 2019). Therefore, this study focused on achieving the dehydration of fructose to HMF and oxidation of HMF to DFF in the same reaction system, realizing a one-pot synthesis of DFF from fructose. This approach not only simplifies the process flow but also minimizes waste generation and solvent use, further enhancing the environmental friendliness of the production method.
The dehydration of fructose to HMF is an acid-catalyzed reaction, whereas the conversion of HMF to DFF involves an oxidation process. Currently, there are two strategies for integrating these reactions within a single reactor. The first approach employs a combination of an acidic catalyst for the dehydration process and a redox catalyst for the oxidation of HMF, requiring the initial transformation of fructose to HMF under the acidic catalyst. Subsequently, an oxidation catalyst is introduced into the reaction system along with oxygen gas or other oxidants to produce DFF. This method, while relatively straightforward in implementation, faces challenges in separating the two catalysts after the reaction, often necessitating the design of one of the catalysts to be magnetic for ease of magnetic recovery and separation (Wei et al. 2021). Moreover, the sequential addition of catalysts adds complexity to the operation, a step that is generally unavoidable to prevent the oxidative degradation of fructose and low yield of DFF if both catalysts were added at the beginning of the reaction. In contrast, the second strategy, which involves the design of a bifunctional catalyst capable of catalyzing both the dehydration of fructose and the oxidation of HMF, appears more streamlined and efficient. A single catalyst that facilitates both reactions can eliminate the issues mentioned above, eliminating the need for separate catalyst recovery steps, and simplifying the overall process. However, the design and preparation of such a bifunctional catalyst was the key. Currently, successful examples of bifunctional catalysts include FeCuPOx, PMA–MIL-101, MoOx/CS, Mo-HNC, CsMVP-HPA, etc. (Liu et al. 2014a; Zhao et al. 2017, 2018a; Zhou et al. 2018; Jia et al. 2021; Nguyen et al. 2024). Among them, heteropolyacids (HPAs) have gained considerable attention in the construction of such bifunctional catalysts due to their abundant protonic acid sites and metal oxide-like structures. To simplify the preparation of bifunctional catalysts, the authors’ idea is to immobilize phosphomolybdic acid (PMA) onto a support, thereby forming a stable structured bifunctional catalyst. In this study, the hydrogen bonding between PMA and chitosan was utilized to form a cross-linked structure, and then a highly dispersed PMA-loaded bifunctional catalyst was obtained through calcination. This catalyst was applied in the one-pot synthesis of DFF from fructose, and achieved excellent results, offering a more straightforward, efficient, and environmentally friendly approach to the synthesis of valuable chemical intermediates from biomass-derived materials.
EXPERIMENTAL
Materials
Fructose, 5-hydroxymethylfurfural (HMF), 2,5-diformylfuran (DFF), phospho-molybdic acid (PMA), ethanol, dimethylsulfoxide (DMSO), and other materials used in the work were obtained in pure form Sigma-Aldrich (Hong Kong S.A.R. China) and were used directly without any modification.
Catalyst Synthesis
Phosphomolybdic acid/chitosan (PMAC) catalysts were synthesized by a facile method. Briefly, 2.0 g of chitosan was dissolved in 100 mL of 2% acetic acid (w/v) to prepare a solution “A”. Subsequently, 0.05 g of phosphomolybdic acid was dispersed into 5 mL of H2O to form solution “B”. Afterward, solution “B” was added dropwise into solution “A” under a continuous and controlled temperature of 80 °C until entirely evaporated. The obtained solid material was collected and further dried overnight in a dry oven at 80 °C. Finally, the dried materials were calcined in a muffle furnace at 350 °C for 10 h to obtain PMAC-1. Likewise, PMAC-2, PMAC-3, and PMAC-4 were prepared through the same procedure with different loadings of PMA (PMAC-x, x = 0.1, 0.3, 0.5 g), respectively. In addition, chitosan-derived char (CN) was prepared using chitosan without PMA following the same procedure. The schematic representation of the prepared catalyst is shown in Scheme 1 below.
Scheme 1. Schematic representation of the preparation process of bifunctional catalyst derived from phosphomolybdic acid and chitosan
Catalytic Reaction
In a typical dehydration reaction, fructose (200 mg) and 30 mg of the catalyst were added to DMSO (5 mL). The mixture was then poured into a round-bottom flask equipped with a reflex condenser, which was placed in a thermostatic oil bath under controlled temperature (90 °C) and stirring (500 rpm) for the desired reaction time (0.5 to 2 h). After the reaction, samples were taken at a specified time from the reaction mixture for quantification of reaction products.
HMF oxidation to DFF
In a typical oxidation reaction of HMF to DFF, 63 mg of HMF, and 30 mg of the catalyst were added to DMSO (5 mL). The mixture was then poured into a 3-necked round bottom flask equipped with a reflex condenser and was placed in a thermostatic oil bath under a controlled temperature of 140 °C and a constant oxygen flow of 10 mL/min was maintained for a specific time. Samples were taken from the reaction mixture at the desired time (1, 2, and 6 h) for product analysis.
One pot and one-step conversion of fructose to DFF
In a one-pot and single-step conversion of fructose DFF, 63 mg of HMF, and 30 mg of the catalyst were added to DMSO (5 mL). The mixture was then poured into a 3-necked round bottom flask equipped with a reflex condenser and was placed in a thermostatic oil bath under a controlled temperature of 130 °C and a constant oxygen flow of 10 mL/min was maintained for a specific time. Samples were taken from the reaction mixture at the desired time (1, 2, and 6 h) for product analysis.
Products Analysis
After the reaction, the solids were removed via centrifugation, filtered with a 0.25 µm syringe filter, and analyzed by high-performance liquid chromatography (HPLC) equipped with a refractive index detector (RID detector Agilent Technologies, 1260 Infinity II). For the analysis of fructose, HMF, and DFF, an organic acid detection column HPX-87H Bio-Rad was used under isocratic conditions. The mobile phase was 5 mM H2SO4 aqueous solution at a flow rate of 0.6 mL/ min and the column temperature was set to 60 °C.
The fructose conversion, HMF yield, DFF yield, HMF conversion, and DFF yield from HMF were calculated using the following Eqs. 1 through 5, respectively,
where [F]0 and [HMF]0 represent the initial concentration of fructose and HMF as substrate at the beginning of the reaction, while [F]t, [HMF]t, and [DFF]t represent the remaining concentration of fructose, HMF, and DFF after the reaction time of t, respectively.
Material Characterization
The structural characterization of all prepared catalysts was completed using Field Emission Scanning Electron Microscopy coupled with Energy Dispersive X-ray analysis (FESEM/EDX, LEO 1530 Field Emission SEM (ZEISS, Germany) and transmission electron microscopy (TEM) using FEI Tecnai G2 F20 S-Twin (FEI Company, Hillsboro, OR, USA). Surface composition and different functional groups of all the catalysts were confirmed through Fourier-transform infrared spectroscopy (FTIR, Nicolet iS 5 FTIR spectrometer, Natick, MA, USA), and Raman spectra for the catalyst were recorded using a ThermoFisher DXR2x instrument (Waltham, MA, USA).
RESULTS AND DISCUSSION
Catalyst Characterization
The synthesis process of PMAC catalyst consists of two stages. The first stage involves the crosslinking of PMA and chitosan in a solution to form a complex. The second stage involves the drying and calcination of this complex to form a stable structure. The process in the first stage aids in the dispersion of PMA, thereby enhancing its dispersibility in the final catalyst and enabling more effective utilization of its catalytic active sites. The structural changes of PMA and chitosan mainly occurred in the second stage, where PMA underwent dehydration due to calcination, while chitosan underwent carbonization. The morphology of the catalyst, as depicted in Fig. 1a through 1f using FESEM images, revealed that the carbon material produced by chitosan after calcination exhibited an irregular block shape, with a relatively flat and clean surface. Upon combination with PMA and subsequent calcination, the surface became rough, exhibiting distinct cracks and the presence of numerous particles. There were no obvious changes in the morphologies of the PMAC catalyst before and after use. Figure 1g through 1i shows the TEM images of CN, PMAC-3, and the used PMAC-3. The CN exhibited a layered structure, PMAC-3, and the used PMAC-3 showed a similar structure, where no aggregated PMA particles were observed. This is further supported by EDS data (Fig. 1j through 1o), which showed a homogeneous distribution of Mo and P within the catalyst particles. These results demonstrated the effectiveness of the synthesis strategy involving the crosslinking of PMA and chitosan followed by calcination in controlling the dispersibility of PMA.
Figure 2a displays the XRD patterns of the catalyst. Characteristically, polymers with crystalline regions can diffract high-intensity sharp peaks, whereas the X-ray peaks for amorphous polymers are much broader. It can be seen in Fig. 2a that the carbon material derived from chitosan revealed a broad peak centered around 20 to 30° in the 2θ range, indicating the presence of an amorphous carbon structure (Kalaiselvimary and Prabhu 2019; Kong et al. 2012). The peak corresponding to chitosan at 25º exhibited reduced intensity and shifted towards 30° in the XRD spectrum of the PMAC samples, suggesting that the introduction of PMA influenced the arrangement of carbon atoms (Tong et al. 2017: Yu et al. 2014; Sadjadi et al. 2023). Moreover, the distinguished peaks at 2θ values of ~10º and ~25º were only present in PMAC samples indicating the existence of PMA (Ahmadian and Anbia 2023).
Fig. 1. FESEM images of catalysts (a) CN, (b) PMAC-1, (c) PMAC-2, (d) PMAC-3, (e) PMAC-4, (f) used PMAC-3, and TEM images of catalyst (g) CN, (h) PMAC-3, (i) used PMAC-3, and EDS of catalyst (j to o) PMSC-3
Fig. 2. (a) XRD patterns, (b) FTIR patterns, and (c) Raman spectrum of CN and PMAC catalysts
Further, Fig. 2b shows the similarities and differences in the FTIR pattern of the samples. All the samples showed three peaks at 1620, 1400, and 1060 cm-1, respectively. The absorbance peak at 1620 cm-1 corresponds to the stretching vibrations of C=O bonds, indicating the presence of carbonyl functional groups in the samples. The peak located at 1400 cm-1 was associated with the bending vibrations of the methyl or methylene groups The absorbance peak at 1060 cm-1 can be attributed to the stretching vibrations of C-O bonds, indicating the presence of oxygen-containing functional groups in the samples. These functional groups were from the chitosan-derived char (CN). In addition to these two peaks, the PMAC samples showed some unique peaks that belong to the PMA species. According to the results reported in the literature, the three absorption peaks at 960, 870, and 796 cm-1 are assigned to stretching vibrations (ν (P-O), ν (Mo-Ot), ν (Mo-Oc-Mo), ν (Mo-Oe-Mo), where Ot, Oc, and Oe refer to terminal, corner, and edge oxygen (Mishra et al. 2010; Santamaria et al. 2017). The XRD and FTIR results confirmed the composite structure of the PMAC catalysts.
The Raman spectra of CN and PMAC-3 are shown in Fig. 2c. Both curves showed Raman peaks at around 1320 and 1600 cm-1, which were attributed to the D band and G band, respectively, of the support material (Lei et al. 2022). For the PMAC-3 curve, an extra absorption peak around 1004 cm-1 was observed, which was ascribed to the stretching vibrations of (MoO) which also revealed that PMA was successfully embedded in chitosan (Baibarac et al. 2011).
Catalytic Dehydration of Fructose to HMF
The synthesized catalysts were initially screened for their catalytic activity of fructose dehydration to HMF. Herein, DMSO was used as the solvent because of its innate behavior of promoting strong acid sites, which could be effective in converting fructose to HMF (Laugel et al. 2014; Tomer and Biswas 2022). Moreover, propagation towards side reactions and humins formation in DMSO is more suppressed, making it an ideal solvent for the dehydration of fructose and stabilization of HMF (Qi et al. 2008; Mushrif et al. 2012). The conversion of fructose to HMF concerning reaction time and different PMA loadings was studied by all synthesized catalysts. Figure 3 depicts the results of fructose conversion to HMF over different catalysts. Notable HMF yields of 41.7%, 45.4%, 68.7%, and 66.7% were observed even at a short reaction time of 1 h over PMAC-1, PMAC-2, PMAC-3, and PMAC-4, respectively. Meanwhile, HMF yield was only 4.5% when CN was used as the catalyst under the same reaction conditions. Similar findings were reported by Chhabra et al. (2023), who found that increasing the PMA loadings in boron nitride (BN) resulted in an 8% increase in HMF yield, whereas using only BN as a catalyst resulted in a significantly lower yield. Likewise, Saghandali et al. (2023) obtained an increased HMF yield of 99.5% from fructose in 45 minutes in DMSO solvent by employing halloysite-supported silicotungstic acid as a catalyst. Both fructose conversion and HMF yield increased with PMA loadings in chitosan suggesting that more acidic sites were introduced to the CN due to increased PMA, which played a crucial role in the fructose dehydration reaction (Chhabra et al. 2023). Considering the slight decrease in HMF selectivity from PMAC-3 to PMAC-4, the effects were probably due to the further hydrolysis or degradation of HMF. These findings are consistent with previous research. Wang et al. (2014) reported that raising the catalyst dosage from 8 to 10 mg reduced HMF yield by 2%, and a further decrease in HMF by 7% was seen when the catalyst dose was increased to 20 mg. The high amounts of catalyst can efficiently cause the degradation and polymerization of HMF to produce byproducts and humins. Thus, optimum loading is critical for achieving efficient product yields. Based on the results, PMAC-3 seems to be the optimal catalyst for fructose to HMF transformation (Zhou et al. 2015).
Fig. 3. Catalytic dehydration of fructose to HMF over CN and PMAC catalyst; Reaction conditions: 30 mg catalyst, 200 mg fructose, 5 mL DMSO, 90 °C, and 1 h
Catalytic HMF Oxidation to DFF
Heteropolyacid (HPA), such as Keggin-type phosphomolybdate acid (HPMo, H3PMo12O40), has been widely explored as a promising catalyst for oxidation reactions (Lei et al. 2021; Zhou et al. 2022; Ping et al. 2023). Such materials are mostly composed of numerous oxygen species in different coordination, including tetrahedral oxygen, bridging oxygen, and terminal oxygen.
Fig. 4. Catalytic oxidation of HMF into DFF over CN and PMAC catalysts; Reaction conditions: 30 mg catalyst, 63 mg HMF, 5 mL DMSO, 140 °C, and 4 h
The shuffling of terminal oxygen species may induce the formation of oxygen vacancies and low-coordination cations (Mo5+ or MO+6) in the lattice, which might be effective in catalyzing oxidation reactions (Xiao et al. 2013; Chang et al. 2020). Recently, Keggin-type heteropolyacid has been explored as a potential catalyst for HMF oxidation to DFF (Zhao et al. 2017; Liu et al. 2019; Wang et al. 2021). Based on this evidence, the role of PMAC in the aerobic oxidation of HMF to DFF was investigated, and the results are shown in Fig. 4. The CN showed limited catalytic activity towards HMF oxidation. After loading with PMA, the PMAC showed considerably improved catalytic activity. With PMAC-1 acting as a catalyst, a 39.4% DFF yield and a 42.2% HMF conversion were attained; moreover, both the DFF yield and HMF conversion rose with increased PMA loading onto chitosan. When PMAC-4 was used as a catalyst, the DFF yield reached a maximum of 73.4% with 80.8% HMF conversion at 140 °C in 4 h. Nevertheless, it is worth mentioning that the DFF selectivity was slightly compromised by introducing higher PMA loadings to the catalyst, and above 90%, DFF selectivity was achievable with all PMAC catalysts. The results demonstrated that high loading of PMA, as it was in PMAC-3 and PMAC-4, was necessary for the high catalytic activity.
One-pot Synthesis of DFF from Fructose
After systematic evaluation of the PMAC catalysts for subsequent fructose dehydration and HMF oxidation, the catalysts were further evaluated as bifunctional catalysts for “one-pot, one-step” direct conversion of fructose to DFF under specified reaction conditions. The experimental results are depicted in Fig. 5.