Elsevier

Applied Catalysis B: Environmental

Volume 235, 5 November 2018, Pages 130-142
Applied Catalysis B: Environmental

Zeolite–supported nickel phyllosilicate catalyst for Csingle bondO hydrogenolysis of cyclic ethers and polyols

https://doi.org/10.1016/j.apcatb.2018.04.053Get rights and content

Highlights

  • Zeolite–supported nickel nanoparticles have been prepared by deposition–precipitation method, forming Ni phyllosilicate precursor.

  • Medium–pore zeolites (ZSM–5 and ZSM–11) exhibit the highest reaction rate in the ring–opening of biomass–derived cyclic ethers and polyols.

  • Up to 38% 1,5–pentanediol selectivity is obtained at low conversion levels over Ni-zeolite catalysts.

  • ZSM–5 and ZSM–11 frameworks have pore confinement effect on cyclo–dehydration of pentane polyols, yielding up to 64% tetrahydropyran.

  • Up to 51% selectivity towards 1,6–hexanediol is achieved during tetrahydropyran–2–methanol conversion over Ni/ZSM–5 catalysts.

Abstract

In this study, a series of zeolites, including small pore (SAPO–34), medium pore (Ferrierite, ZSM–5, and ZSM–11), and large pore (MOR and BETA) zeolites were explored as supports for nickel catalysts. The catalysts were prepared by the deposition–precipitation (D–P) method to obtain ca. 10 wt% Ni loading. The structures of prepared catalysts were elucidated using a wide range of characterization techniques such as BET, XRD, TEM, H2–TPR, NH3–TPD, FTIR, and XPS. It was found that 1:1 nickel phyllosilicate structures is formed over D–P nickel catalysts, while 2:1 nickel phyllosilicate was identified only over Ni/BETA catalyst. Furthermore, Ni/ZSM–5 catalyst was also prepared by impregnation method as a reference to investigate the effect of preparation method. The impact of the topological structures of zeolite on the catalytic activity of reactants (i.e., cyclic ethers derived from furanic compounds), reaction intermediates (i.e., polyols) was subsequently investigated. The ring–opening of THFA can proceed over Ni/Ni phyllosilicate species on the external surface of the catalyst, producing pentane polyols. Then, the pentane polyol intermediates can further undergo ring–closure pathway, predominantly over the internal acid sites within zeolite structures (except for Ni/SAPO–34). It was found that medium–pore zeolites (ZSM–5 and ZSM–11) with mild acidity exhibit the highest reaction rate with a THP yield up to ca. 64%. Similarly, the highest catalytic activity during THPM conversion was achieved over ZSM–5– and ZSM–11– supported nickel catalysts, with up to 51% selectivity towards 16HDO. Comparing the zeolite maximum pore sizes and the kinetic diameters of the reactants and products with the catalytic performance, it is clear that topological structures of ZSM–5 and ZSM–11 have pore confinement effect on the reaction mechanism and catalytic activity in the hydrogenolysis of cyclic ethers and polyols.

Introduction

The dependence of the chemical industry on traditional oil feedstocks is of growing concern, especially when oil reserves are continually being depleted. Furthermore, the environmental hazards that the usage of oil poses have become a strong motivation for the development of environmentally friendlier, bio–renewable feedstock for the chemical industry [1]. Furanic compounds and their derivatives have been well established as a bio–renewable feedstock for the production of a wide range of high value–added chemicals [2]. As a well–known example, furfural can be obtained from acid–catalyzed dehydration of pentoses obtained from plant biomass [3,4]. Tetrahydrofurfuryl alcohol (THFA), which can be produced from the total hydrogenation of furfural over Ni–based catalysts [5,6], can then undergo ring opening to form 1,5–Pentanediol (15PDO) when the ring–opening is catalyzed by noble metals such as Rh or Ir modified by Mo or Re oxide [[7], [8], [9], [10]]. Similarly, 5–hydroxymethyl furfural (HMF) can undergo sequential hydrogenation and ring–opening to produce value–added 1,6–hexanediol (16HDO) [11]. These linear carbon chains with both terminal OH groups, known as α,ω–diols, are widely used as monomers for the production of polyesters and polyurethanes.

For such ring–opening reactions, catalysis by noble metals have been extensively studied. However, due to the rarity and price of such metals, it is desired to use non–noble metals as potential replacements for these catalysts. A pioneer work developed a multiple approach to synthesize 15PDO from THFA via a dehydration–hydration–hydrogenation pathway, with an overall yield of ca. 70% [12]. Very recently, this process has been modified by Dumesic et al. for the production of 15PDO from biomass–derived THFA over inexpensive catalysts, reaching a significantly higher overall yield of 90% [13]. The process consists of the following three steps: (1) dehydration of THFA to 3,4–2H–dihydropyran (DHP) over γ–Al2O3 in the vapor phase; (2) hydration of DHP to 2–hydroxytetrahydropyran (2–HTHP) without addition of any external catalyst in the liquid phase; and (3) ring–opening tautomerization of 2–HTHP to δ–hydroxyvaleraldehyde which can be subsequently hydrogenated into 15PDO over Ru catalysts. Similar multi–step approach was applied to ring–opening of tetrahydropyran–2–methanol (THPM) into 16HDO, resulting in a maximum 34% overall yield of 16HDO [14].

Our previous study [15] has demonstrated that Ni itself is capable of ring opening of THFA at the more sterically hindered secondary Csingle bondO bond, yielding mostly 15PDO along with 1,2,5–pentanetriol (125PTO), which was identified for the first time in aqueous–phase hydrogenolysis of THFA at high temperatures. Furthermore, nickel supported on ZSM–5 is a promising candidate as an inexpensive catalyst with high chemoselectivity for both ring–opening and ring–closure, yielding tetrahydropyran (THP). THP has a wide range of applications as a solvent or as an intermediate in organic synthesis to produce glutaric acid, 1,5–dichloropentane, heptanediamine, and pimelic acid [16,17]. Additionally, it can undergo ring–opening/dehydration to produce pentadienes [18]. The key advantage of this route is the use of a relatively cheap non–noble metal such as Ni in the direct synthesis of THP from THFA. Therefore, it is important to further explore the factors affecting the performance of such Ni–based catalysts.

One of the most critical factors affecting catalyst activity is the catalyst preparation method [19,20]. nickel–zeolite catalysts are typically prepared by the relatively simple incipient wetness impregnation (IWI) method. However, catalysts prepared by impregnation methods typically suffer from relatively low metal dispersion and hence low catalytic activity [21]. To improve the dispersion and activity of the nickel–zeolite catalysts, the ion exchange and deposition–precipitation (D–P) method [[22], [23], [24], [25]] were studied and it was found that those methods are able to boost higher initial activity and recyclability after repeated runs as compared to impregnation, particularly when using organic solvents [23]. One key reason for this difference is the improved degree of dispersion of nickel on the support surface for the latter two methods as compared to the impregnation one. However, ion exchange method is limited to relatively low metal loading. Previous studies also reported the presence of different nickel species, including nickel hydroxide, nickel oxide, nickel phyllosilicate species, in zeolite–supported catalyst prepared by D–P method [[21], [22], [23],26]. Besides, our previous studies showed that the nickel catalysts prepared via phyllosilicate precursor could suppress sintering of nickel particles [27]. Hence, it is of great interest to unravel if this stabilizing effect of highly dispersed nickel using the D–P method leads to a good catalytic performance during ring–opening reactions in liquid (aqueous) phase.

Besides preparation methods, another important factor in catalyst activity and effectiveness is the accessibility to the active sites and the ability of reactants/products to diffuse in/out the pores to reach the active sites. Since the structure of the support directly affects the diffusivity of reactants and products in and out of the pores, the structure of the zeolite may also affect the yield and selectivity of the reaction. The structure also affects the shape selectivity of the catalyst [28], which means that the structure may improve selectivity towards given products. Indeed, the effect of the zeolite structure has already been reported in previous studies using bare zeolites [29] and zeolite–supported Ni catalysts [30,31].

In order to get a better understanding of the physicochemical properties of zeolites on ring–opening reactions, this paper aims to understand the effect of catalyst preparation method (D–P) and zeolite structure of zeolite–supported Ni catalysts on performance and recyclability. To achieve this, various zeolites, including small pore zeolites (SAPO–34), medium pore zeolites (Ferrierite, ZSM–5, and ZSM–11), and large pore zeolites (Mordenite and BETA) were investigated for the hydrogenolysis of cyclic ethers and polyols derived from furanic compounds.

Section snippets

Catalyst preparation

The ammonium form of Ferrierite (FER), ZSM–5, Mordenite (MOR) and BETA zeolites with SiO2/Al2O3 ratios of 20, 30, 20, and 25, respectively, were obtained from Zeolyst and used as catalyst support. Nickel (II) Nitrate hexahydrate (98%, Alfa Aesar), Urea (ACS, 99.0–100.5% from Alfa Aesar) were used for catalyst preparation. THFA (99 wt%, Sigma Aldrich), THPM (98%, TCI), 1,2,5–pentanetriol (97%, TCI), 1,5–pentanediol (97%, Sigma Aldrich), 1,2,6–hexanetriol (96 wt%, Sigma Aldrich), and

Characterization results

The physicochemical properties for the parent zeolites and Ni–zeolite catalysts are summarized in Table S2 (Supplementary Material) and Table 2, respectively. The Ni loading of the catalysts was fixed at about 10 wt%. The nickel loadings determined by XRF range from 9.2 wt% to 10.3 wt%, being the differences attributed to the differences in catalyst supports and preparation methods. The variation of the solution pH during D–P of Ni (II) species on ZSM–5 is shown in Fig. S1 (Supplementary

Formation of Ni phyllosilicate precursors during D–P process

As presented in the result section, various characterization techniques were employed to elucidate the nature of active sites on the surface of prepared Ni–zeolite catalytic systems. Compared to the bare zeolites, all Ni–zeolites catalysts present larger external surface areas, total pore volumes, and mesoporous volumes, but smaller values of BET surface area, microporous surface area and pore volume. This is ascribed to the fact that the deposited Ni species partially block the micropores of

Conclusions

The physicochemical and structural properties of different zeolites, including small pore (SAPO–34), medium pore (Ferrierite, ZSM–5, and ZSM–11), and large pore zeolites (Mordenite and BETA), were investigated in C–O hydrogenolysis of cyclic ethers (THFA and THPM) and polyols using Ni catalysts. nickel–zeolite catalysts were synthesized via deposition–precipitation (D–P) method and analyzed with a wide range of characterization techniques. The characterization results show the existence of

Acknowledgements

The authors gratefully thank the Singapore Agency for Science, Technology and Research (A*STAR), National University of Singapore, and NEA (ETRP 1501 103) for their generous financial supports. The help of Dr. Liu Dapeng with synthesizing ZSM–11 and SAPO–34 zeolites is also greatly acknowledged. One of us (E.S.) is grateful to A*STAR for the Singapore International Graduate Award (SINGA) scholarship.

References (49)

  • C. Wang et al.

    High quality diesel-range alkanes production via a single-step hydrotreatment of vegetable oil over Ni/zeolite catalyst

    Catal. Today

    (2014)
  • A. Lugstein et al.

    Hydroisomerization and cracking of n-octane and C8 isomers on Ni-containing zeolites

    Appl. Catal., A

    (1999)
  • D. Liu et al.

    Catalytic conversion of ethanol over ZSM-11 based catalysts

    Appl. Catal. A

    (2016)
  • S. Tomiyama et al.

    Preparation of Ni/SiO2 catalyst with high thermal stability for CO2-reforming of CH4

    Appl. Catal. A

    (2003)
  • C. Zhao et al.

    Understanding the impact of aluminum oxide binder on Ni/HZSM-5 for phenol hydrodeoxygenation

    Appl. Catal. B

    (2013)
  • M. Tamura et al.

    Comprehensive IR study on acid/base properties of metal oxides

    Appl. Catal. A

    (2012)
  • T. Lehmann et al.

    Physico-chemical characterization of Ni/MCM-41 synthesized by a template ion exchange approach

    Microporous Mesoporous Mater.

    (2012)
  • S. Sitthisa et al.

    Kinetics and mechanism of hydrogenation of furfural on Cu/SiO2 catalysts

    J. Catal.

    (2011)
  • S. García-Fernández et al.

    The role of tungsten oxide in the selective hydrogenolysis of glycerol to 1, 3-propanediol over Pt/WOx/Al2O3

    Appl. Catal. B

    (2017)
  • M.S. Kumar et al.

    Evidence of the vital role of the pore network on various catalytic conversions of N2O over Fe-silicalite and Fe-SBA-15 with the same iron constitution

    Appl. Catal. B

    (2006)
  • T. Garetto et al.

    The origin of the enhanced activity of Pt/zeolites for combustion of C2–C4 alkanes

    Appl. Catal. B

    (2007)
  • S.K. Brar et al.

    Platform Chemical Biorefinery: Future Green Chemistry

    (2016)
  • Y. Nakagawa et al.

    Catalytic reduction of biomass-derived furanic compounds with hydrogen

    ACS Catal.

    (2013)
  • X. Li et al.

    Furfural: a promising platform compound for sustainable production of C4 and C5 chemicals

    ACS Catal.

    (2016)
  • Cited by (67)

    • Steam reforming of bio-alcohols over Ni-M (Cu, Co, Pt)/MCF-S (MgO, La<inf>2</inf>O<inf>3</inf>, CeO<inf>2</inf>) for renewable and selective hydrogen production: Synergistic effect of MCF silica and basic oxides on activity and stability profiles

      2023, Catalysis Today
      Citation Excerpt :

      In the case of Ni-Cu/MCF-CeO2 [Fig. S8 (b)], Cu-Ni-Co/MCF-MgO [Fig. S8 (c)], and Pt-Ni-Cu/MCF-La2O3 [Fig. S8 (d)] catalysts, Ni2+ peaks of Ni 2p1/2 are seen into single satellite peak at BE of 880.9 eV, 879.7 eV, and 884.3 eV, respectively, and Ni 2p3/2 peaks are identified into two broad satellite peaks at BEs of 862.8 and 868.9 eV, 862.5 and 868.2 eV and 866.3 and 873 eV, respectively through spin-orbit splitting energy of 17.7 eV [70,75]. Likewise, as described before, due to calcination, decomposition of Ni(OH)2 into NiO and H2O takes place, which is identified as a multiplet-split Ni 2p3/2 and Ni 2p1/2 peaks of NiO [70,76]. For Ni-Cu/MCF (72 h) catalyst [Fig. S8 (e)], we did not observe any Ni 2p spectra which could be because of the excessive dispersion of active Ni nanoparticles into the highest pore diameter of MCF (72 h) support (22.2 nm).

    View all citing articles on Scopus
    View full text