Elsevier

Catalysis Today

Volume 379, 1 November 2021, Pages 70-79
Catalysis Today

Mild-Temperature hydrodeoxygenation of vanillin a typical bio-oil model compound to Creosol a potential future biofuel

https://doi.org/10.1016/j.cattod.2020.05.066Get rights and content

Highlights

  • Solvent polarity index and hydrogen uptake influence conversion and selectivity.

  • Pd/C and PdRh/Al2O3 showed excellent stability after three cycles of experiments.

  • Orders of effect of catalyst and process parameters are summarised as follows.

  • Creosol Selectivity: PdRh/Al2O3 >> Pd/C >> Rh/Al2O3 >> Pt/SiO2 > Pd/Al2O3 > Pt/C.

  • Degree of deoxygenation: catalyst loading >> temperature >> pressure >> agitation.

Abstract

This study reports mild temperature hydrodeoxygenation (HDO) of vanillin an oxygenated phenolic compound found in bio-oil to creosol. It investigates the sensitivity of vanillin HDO reaction to changes in solvent, catalyst support and active metal type, and processing parameters using 100 ml batch reactor. The processing parameters considered include temperature (318 K – 338 K), hydrogen gas pressure (1 MPa – 3 MPa), catalyst loading (0.1 kg/m3 – 0.5 kg/m3), and agitation speed (500 rpm–900 rpm). As expected, significant variation in conversion and product selectivity was displayed in the results. Among the solvents considered, 2-propanol and ethyl acetate produced the best performance with conversion close to 100 % and selectivity toward creosol above 90 %. Remarkable differences were found in the H2 uptake during VL HDO reaction under different catalyst. The hierarchy in H2 uptake of the catalysts include: Pd/C > PdRh/Al2O3 > Pd/Al2O3 = Pt/C > Pt/SiO2 >> Rh/Al2O3. This was correlated to catalytic performance; Pd/C emerged as the best among the monometallic catalysts with 71 % selectivity toward creosol, but consumed 9 mmol of hydrogen per mol of vanillin converted. While the prepared bimetallic PdRh/Al2O3 catalyst consumed slightly lower amount of hydrogen (8 mmol), and produced significantly higher selectivity toward creosol (99 %). Even after three cycles the prepared catalyst demonstrated superior performance over the monometallic catalysts with selectivity toward creosol above 80 %. The reaction condition that maximises the degree of deoxygenation to creosol derived via Taguchi analysis includes temperature 338 K, hydrogen gas partial pressure 3.0 MPa, catalyst loading 0.5 kg/m3, and agitation speed 500 rpm.

Introduction

Fossil-derived fuels such as petroleum, coal and natural gas have long been the major source of energy to the world [1]. However, the capacity of light crude oil reserves responsible for producing most hydrocarbon fuels have been declining for many years. In addition, consumption of fossil derived fuels adversely affects the environment. As a result of these factors, attention has shifted toward alternative sources of energy. Preferably sources that are renewable and sustainable should be exploited [[2], [3], [4], [5]]. Plant biomass represents a promising source of energy because of the renewability and reduced carbon footprint [6]. Liquid fuel known as bio-oil is usually produced from plant biomass via fast pyrolysis (FP), this process occurs rapidly (typically 1–2 s) at temperatures around 773 K in the absence of air [[7], [8], [9], [10], [11]]. As a result, produced bio-oil usually contains a significant amount of thermodynamically unstable oxygenated compounds [[12], [13], [14], [15]]. Up to 400 different oxygenated compounds have been reported in analytical studies on bio-oil [[16], [17], [18]]. In contrast, the oxygen content in conventional fuels is less than 1 wt %. This high oxygen content in bio-oils (20−30 wt %) leads to problems such as blocked filters, excessive corrosion, pump breakdown etc. during initial attempts to substitute conventional fuels with bio-oils in operation of furnaces, gas turbines, boilers and space heaters [[19], [20], [21], [22], [23]]. Hence, upgrading is necessary for bio-oils to fulfil their potential as substitute fuels. Hydrodeoxygenation (HDO) is the leading technology for upgrading them. It involves rejection of oxygen from the bio-oil with hydrogen as the reducing agent, and usually involves temperatures of 423–573 K and 7.5–30 MPa hydrogen gas pressure. However, the complexity of bio-oil and HDO reaction network compelled many laboratory scale studies to single or mixtures of model compounds present in bio-oil [[24], [25], [26], [27], [28]]. Common model compounds used in past studies on HDO include cresol, guaiacol, and anisole. These compounds were chosen because they contain multiple functional groups and are member of the guaiacyl species [29]. These species are favoured as model of oxygenates in bio-oil because they represent the primary structure of lignin fraction used to produce bio-oil. Vanillin (VL), an aromatic aldehyde compound cited in different analytical studies on bio-oil, is of interest in this work. This compound is selected because it contains a carbonyl group which is largely responsible for the thermodynamic instability of real bio-oil [[30], [31], [32], [33], [34]]. Moreover, the three functional groups aldehyde, ether, and hydroxyl present in the structure of vanillin makes it a good representative model of oxygenates present in bio-oil. Finally, it belongs to the same group as cresol, guaiacol, and anisole. Catalysts used in past studies on real bio-oil and model compound HDO reactions include transition sulfided metals (CoMoS, and NiMoS), noble or precious metals (Pd, Pt, Ru and Rh) and non-noble metals (Ni, Cu, Fe, and their alloys) [28,35]. Looking at the performance of these catalysts in past HDO studies, undoubtedly noble metals demonstrated superior activity. However, the exorbitant cost in procuring these metals and ease at which they are poisoned raises concerns about the economics of the process [[36], [37], [38]]. Fortunately, resistance of these metals to poison can be increased by combining them with another metal. In addition, higher selectivity toward deoxygenated products was achieved in past HDO studies that utilised bimetallic catalysts [[39], [40], [41]]. Hence, a novel bimetallic catalyst comprising of palladium and rhodium on alumina support was synthesised, characterised and tested in this work. Furthermore, the catalytic performance of noble metals (Pd, Rh and Pt) on inert (c) and conventional supports (SiO2 and Al2O3) were compared to the synthesised bimetallic catalyst. The role of solvent and processing parameters in VL HDO reaction was probed using different solvents and conditions. Finally, Taguchi analysis was used to derive the best mild condition for transforming vanillin to creosol which is a potential future biofuel.

Section snippets

Synthesis of PdRh/Al2O3 catalyst

PdCl2 solution (10.0 g, 5 wt % in 10 wt % HCl, Sigma – Aldrich) was added dropwise to methanol (30 mL, Fischer Scientific, 99 %) under vigorous stirring and then 5 wt % Rh/Al2O3 (10.0 g, Johnson Matthey) was introduced. The resultant mixture was stirred at room temperature for 4 h and dried overnight at 298 K. Further drying in an oven set at 353 K was carried out for 2 h prior to calcination at 773 K for 4 h.

Characterisation of the catalysts

Nitrogen adsorption – desorption isotherms of the catalyst used in this study were

Results and discussion

The range of conditions investigated ensures liquid phase only reaction with insignificant coke and gas formation. Also, negligible changes in the weight of solvents used were observed, which suggests the solvents are not converted into gases. Consequently, only two products vanillyl alcohol and creosol were recorded, meaning the solvents were not involved in the reaction. This section is organized as follows; characteristics of the catalysts used are presented and discussed in section 3.1,

Conclusion

In this work, vanillin (VL) a model compound of bio-oil was used to investigate the effect of solvent, catalyst active metal, supports and processing parameters on hydrodeoxygenation (HDO) reaction. It was observed that non-catalytic and catalytic support influences on the reaction are not significant when compared to the catalytic active metals (Pd, Pt, Rh, and PdRh). Product distribution and conversion changed significantly with the reaction solvents. This was correlated to observed

CRediT authorship contribution statement

Elias Aliu: Methodology, Investigation, Writing - original draft, Visualization. Abarasi Hart: Formal analysis, Writing - review & editing, Supervision. Joseph Wood: Conceptualization, Resources, Supervision, Project administration, Funding acquisition.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The financial support for this work was provided by School of Chemical Engineering, University of Birmingham.

References (53)

  • Q. Lai et al.

    Appl. Catal. A Gen.

    (2016)
  • M. Patel et al.

    Renew. Sustain. Energy Rev.

    (2016)
  • Y. Wang et al.

    Bioresour. Technol.

    (2012)
  • A.V. Bridgwater

    Catal. Today

    (1996)
  • A.V. Bridgwater

    Appl. Catal. A Gen.

    (1994)
  • M.E. Boucher et al.

    Biomass Bioenergy

    (2000)
  • A.V. Bridgwater

    Biomass Bioenergy

    (2012)
  • S. Echeandia et al.

    Fuel

    (2014)
  • S. Boullosa-Eiras et al.

    Catal. Today

    (2014)
  • J. Jae et al.

    J. Catal.

    (2011)
  • L. Busetto et al.

    Fuel

    (2011)
  • C.A. Fisk et al.

    Appl. Catal. A Gen.

    (2009)
  • E. Laurent et al.

    Appl. Catal. A Gen.

    (1994)
  • L. Zhang et al.

    Renew. Sustain. Energy Rev.

    (2013)
  • P.M. Mortensen et al.

    Appl. Catal. A Gen.

    (2011)
  • T. Hosoya et al.

    J. Anal. Appl. Pyrolysis

    (2009)
  • R. Lou et al.

    J. Anal. Appl. Pyrolysis

    (2010)
  • C.A. Mullen et al.

    Fuel Process Technol.

    (2010)
  • Q. Yang et al.

    J. Anal. Appl. Pyrolysis

    (2010)
  • D.C. Elliott

    Curr. Opin. Chem. Eng.

    (2015)
  • A.R. Ardiyanti et al.

    Appl. Catal. A Gen.

    (2012)
  • H. Wan et al.

    J. Catal.

    (2014)
  • I. McManus et al.

    J. Catal.

    (2015)
  • J. He et al.

    J. Catal.

    (2014)
  • B.S. Akpa et al.

    J. Catal.

    (2012)
  • F.H. Mahfud et al.

    J. Mol. Catal. A Chem.

    (2007)
  • Cited by (18)

    View all citing articles on Scopus
    View full text